INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

AN IMPORTANT FUNCTION of the mouse and rat medullary thick ascending limb (MTAL) is to reabsorb luminal HCO₃. The first step of HCO₃ absorption is believed to be mediated virtually completely by apical Na⁺/H⁺ exchange (reviewed in Ref. 13).

In contrast, little is known about the specific pathways involved in HCO₃ transport across the basolateral membrane of MTAL cells. In particular, although recent studies have established that mRNAs transcribed from the AE anion exchanger genes AE1 (31) and AE2 (6) are expressed in the rat MTAL and that AE2 transcript (6) and protein (2) expression are higher in the MTAL than in other nephron segments in rat, the possible roles of these Na⁺-independent Cl/HCO₃ exchangers in mediating transcellular HCO₃ absorption in the MTAL remains largely unknown. In 1986, Hebert (15) first suggested the presence of a basolateral membrane Cl/HCO₃ exchanger by quantitative morphological measurements of cells of the in vitro perfused mouse MTAL. He proposed that this exchanger operates in parallel with a Na⁺/H⁺ antiporter to mediate hypertonic cell volume regulation. However, both of these exchangers were considered to be quiescent under isotonic conditions in the absence of arginine vasopressin (AVP) (15). Subsequent studies in mouse (23) and rat (24) MTAL suspensions have also failed to detect Cl/HCO₃ exchange under isotonic conditions and in the absence of AVP by measuring intracellular pH (pH_i).

We now report a direct study of Cl/HCO₃ exchange across both types of luminal (LMV) and basolateral (BLMV) membrane vesicles isolated from rat MTALs (4) to examine, by ³⁶Cl influx studies, whether Cl/HCO₃ exchange is present in this nephron segment. The results provide direct evidence for the presence of Cl/HCO₃ exchange not only at the basolateral membrane, but also at the luminal membrane. The BLMV and LMV exchange activities, however, could be distinguished from one another on the basis of their sensitivities to DIDS and to pH_i. The ~165-kDa AE2 polypeptide was restricted to BLMV. AE1-related polypeptides were present predominantly in BLMV, as well as at variable, lower levels also in LMV.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Preparation of MTAL tubules. The tubule isolation procedure was similar to that described recently (4). In brief, male Sprague-Dawley rats weighing 250-300 g were anesthetized with pentobarbital sodium. Kidneys were removed quickly, decapsulated, and sliced sagittally. Slices were transferred in fresh iced (0-4°C) Hanks' modified medium containing (in mM) 115 NaCl, 0.4 MgSO₄, 0.5 MgCl₂, 0.4 KH₂PO₄, 0.3 Na₂HPO₄, 25 NaHCO₃, 10 HEPES, 4 KCl, 1.2 CaCl₂, 5 glucose, 5 L-leucine, and 1 mg/ml bovine serum albumin; pH 7.40 (bubbled with 95% O₂-5% CO₂). Under stereomicroscopic control, the inner stripe of the outer medulla, recognized by its reddish color, was carefully separated from each slice by removing completely the outer part of the outer medulla as well as the inner medulla. The resulting tissue was subjected to collagenase treatment as described (4). In the final suspensions, most of the tubules (>95%) proved to be MTAL in origin, based on immunofluorescent staining with Tamm-Horsfall protein (4), a specific marker for the thick ascending limb. We were unable to detect significant activity of maltase, a marker of the proximal tubule brush-border membrane, in either whole homogenates or in the final apical fractions, further indicating that the starting material was not significantly contaminated by tubules from the pars recta.

Isolation of plasma membranes. Typically, the preparation began with ~15 mg protein of MTAL tubules obtained from the kidneys of 10 rats. LMV and BLMV were prepared from purified rat MTAL tubules as recently described in detail (4). We have recently demonstrated that the BLMV and LMV preparations have a right-side-out orientation (5). Compared with homogenate, the basolateral marker activity of Na⁺-K⁺-ATPase was enriched more than 9-fold in the BLMV and only 0.5-fold in the LMV. Immunoblot analysis confirmed comparable enrichment of the ₁-subunit polypeptide of Na⁺-K⁺-ATPase in BLMV compared with LMV (see below). In contrast, the apical marker activity of -glutamyltransferase was enriched >10-fold in the LMV and 2-fold in BLMV (4). Immunoblot analysis of the 31-kDa subunit of the vacuolar H⁺-ATPase (7) also confirmed its enrichment in LMV over BLMV (see below). Transport assays were performed after overnight storage of the vesicles at 85°C.

Transport measurements. ³⁶Cl uptake into the membrane vesicles was assayed at ambient temperature (20-25°C) by a rapid filtration technique. The vesicles were preequilibrated at room temperature for 2 h to load with desired constituents. For each experiment, the specific conditions are given in the legends to Figs. 1-7. In general, a 10-µl aliquot of either BLMV or LMV (10-40 µg protein) was added to 100-300 µl of appropriate reaction medium containing ³⁶Cl (~1 µCi/ml). The reaction was stopped with 1.5-ml ice-cold solution containing 20 mM Tris-HEPES pH 7.40 and the desired potassium gluconate concentration to maintain constant osmolality. This suspension was rapidly filtered on the center of a 0.45-µm prewetted Millipore cellulose filter (HAWP) and washed with an additional 15 ml ice-cold stop solution. In all experiments, vesicle uptake was corrected for nonspecific isotopic binding to the filter. The filters were dissolved in 3 ml of scintillant (Filter-count, Packard), and radioactivity was determined using a -scintillation counter.

Antibodies. Mouse monoclonal antibody a5 to chick ₁-subunit polypeptide of Na⁺-K⁺-ATPase was obtained as hybridoma supernatant from the Developmental Studies Hybridoma Bank (Iowa City, IA). Mouse monoclonal antibody E11 to bovine kidney 31-kDa subunit of vacuolar H⁺-ATPase (7), the gift of S. Gluck, was also used as hybridoma supernatant. Rabbit polyclonal antibody to mouse AE2 amino acids 1224-1237 (2, 6, 18, 36, 32), mouse monoclonal antibody to rat AE1 holoprotein (2), the gift of D. Biemesderfer, and rabbit polyclonal antibody to mouse AE1 amino acids 917-929 (9) have been described previously.

Immunoblot analysis. Freshly fractionated BLMV and LMV (4) were solubilized in SDS-load buffer, incubated with vortexing at 20°C for 30 min, then frozen at 80°C until used. Proteins were electrophoretically separated on 3-20% (linear gradient) polyacrylamide gels, then transferred to nitrocellulose (Schleicher and Schuell, Keene, NH). Immunoblots were performed as described previously (6), with minor modification. Blocked nitrocellulose strips were incubated with diluted primary antibodies, where indicated, in the presence of irrelevant peptide or peptide antigen at 24 µg/ml. After washing and incubation with peroxidase-coupled goat anti-Ig (Jackson Immunoresearch, West Grove, PA), blots were developed with enhanced chemiluminescence reagents from New England Biolabs (Beverly, MA).

Statistical methods. All data are represented as means ± SE. Comparisons among groups were generally carried out by a two-way ANOVA or t-test. For all analyses, statistical significance was defined as P < 0.05. AE-mediated ³⁶Cl influx values determined at pH_i 6.1-8.0 were fit to a four-parameter logistic sigmoid equation using Inplot 3.1 (GraphPad software) as follows: v = A + B  A/[1 + (10^C/10^X)^D], where v = measured ³⁶Cl influx, A = the minimal value for ³⁶Cl influx, B = the maximal value for ³⁶Cl influx, C = pH_{i (50)} (the pH_i at which v is half-maximal), X = pH_i, and D is the Hill coefficient.

Materials. H³⁶Cl obtained from New England Nuclear was titrated to neutrality with tetramethylammonium (TMA) base to produce TMA³⁶Cl. DIDS was obtained from Research Organics (Cleveland, OH).

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

H⁺ and HCO₃ gradient-dependent uptake of chloride in BLMV and LMV. Figure 1 shows the effects of pH and/or HCO₃ gradients on time-dependent ³⁶Cl uptake by basolateral and apical membrane vesicles. For both preparations, an inside alkaline pH gradient [pH_i 7.8, external pH (pH_o) 5.5] induced transient stimulation of Cl uptake, compared with that in the absence of a pH gradient (pH_o = pH_i = 7.8). The presence of an outward HCO₃ gradient of 55 mM:2.62 mM with the same pH gradient induced a further stimulation of Cl uptake in both types of vesicles, with maximal Cl accumulation achieved within ~2 and 6 min, respectively, in BLMV and LMV. The peak Cl uptakes (overshoots), were 3- and 1.5-fold greater than the respective equilibrium values for BLMV (measured at 3 h) and LMV (5 h). In both preparations, however, even the equilibrium levels of intravesicular Cl remained two to five times greater than in the absence of either H⁺ or HCO₃ gradients. These differences in the 3-h and 5-h levels of Cl may suggest that the HCO₃ and H⁺ gradients across the vesicular membranes dissipated very slowly, thus retarding efflux of Cl that had been concentratively transported into the plasma membrane vesicles.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Effect of H+ and HCO3 gradient on Cl uptake by BLMV (A) and LMV (B) isolated simultaneously. Vesicles were preincubated for 2 h with a pH 7.8 medium in the presence of HCO3 [in mM: 100 mannitol, 3 EGTA, 100 N-methyl-D-glucamine (NMG) gluconate, 55 NMG HCO3, and 100 Tris-HEPES gassed with 95% N2-5% CO2] or in the absence of HCO3 (110 mannitol, 3 EGTA, 100 NMG gluconate, and 200 Tris-HEPES gassed with 100% N2). Chloride (2 mM 36Cl) uptake was then assayed by diluting vesicles 1:21 into pH 5.5 buffer containing either (in mM) 100 mannitol, 3 EGTA, 155 NMG gluconate, and 100 Tris-MES gassed with 95% N2-5% CO2, or 100 mannitol, 3 EGTA, 100 NMG gluconate, and 200 Tris-MES gassed with 100% N2. Uptake in the absence of pH or HCO3 gradient was determined by rapid dilution of vesicles into reaction media identical to preincubation media, except for the presence of 36Cl. Values are means of 6 determinations from 2 different BLMV and LMV preparations. When SE bar is not shown, it was smaller than the symbols. pHi, intracellular pH; pHo, external pH.

View this table: [in this window] [in a new window]   Table 1.   Effect of FCCP on the 5-h level of 36Cl uptake in BLMV and LMV

Effect of an electrical potential across the membrane vesicles on ³⁶Cl uptake. We next investigated whether significant Cl-conductive pathways exist in these membranes and whether DIDS, an established inhibitor of Cl/HCO₃ exchangers, would inhibit Cl uptake. These experiments were performed in the absence or presence of transmembrane potentials generated by preincubation of membrane vesicles with valinomycin. Generation of an inside-positive membrane potential by imposition of an inwardly directed K⁺ gradient markedly stimulated ³⁶Cl uptake in BLMV (Fig. 2, left) and in LMV (Fig. 2, right), indicating significant Cl-conductive pathways in both preparations. Figure 2 shows, however, that 2 mM DIDS had no significant effect on ³⁶Cl uptake in BLMV and LMV under both unclamped and voltage-clamped conditions.

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Effect of transmembrane potential difference on chloride uptake by BLMV and LMV. Both types of vesicles were preincubated for 2 h with medium containing (in mM) 110 mannitol, 200 Tris-HEPES (pH 7.8), 3 EGTA, and either 100 potassium gluconate or 100 tetramethylammonium (TMA) gluconate. All vesicles were pretreated for 2 h with valinomycin (10 µg/mg protein). Chloride (2 mM 36Cl) uptake at 9 s was then assayed by rapid dilution of vesicles 1:21 into reaction media containing (in mM) 110 mannitol, 200 Tris-HEPES (pH 7.8), 3 EGTA, 100 potassium gluconate, and either 4 sodium gluconate or 2 DIDS as the disodium salt. Values are means ± SE of 6 determinations from 2 different LMV and BLMV preparations. The inhibitory effects of DIDS on both types of plasma membranes were not statistically significant (ANOVA).

Comparison of the effects of DIDS on Cl/HCO₃ exchange in BLMV and LMV. Figure 3 shows that pH- and HCO₃ gradient-stimulated ³⁶Cl uptakes in BLMV and LMV were sensitive to DIDS inhibition in the micromolar range, suggesting that the stimulation of ³⁶Cl uptake by HCO₃ gradients in BLMV and LMV is direct rather than secondary to generation of an inside-positive membrane potential. Figure 3 also shows that the BLMV and LMV inhibition curves differed significantly (P < 0.05), with IC₅₀ values for DIDS of 3.2 ± 0.9 and 15.2 ± 5.2 µM.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Inhibition of BLMV () and LMV (×) Cl/HCO3 exchangers by DIDS. BLMV and LMV were preincubated for 2 h with a pH 7.8 medium containing (in mM) 100 mannitol, 3 EGTA, 100 TMA gluconate, 55 TMA HCO3, and 100 Tris-HEPES, gassed with 95% N2-5% CO2. Chloride (2 mM 36Cl) uptake at 9 s was then assayed by rapid dilution of vesicles 1:21 into pH 5.5 buffer containing (in mM) 100 mannitol, 3 EGTA, 155 TMA gluconate, and 100 Tris-MES gassed with 95% N2-5% CO2. Sodium gluconate was added as appropriate to maintain the Na+ concentration at 4 mM. Values are means ± SE of 6 determinations from 2 different LMV and BLMV preparations.

pH_i dependence of the BLMV and LMV anion exchangers. Figure 4 compares the effect of pH_i on pH- and HCO₃ gradient-stimulated ³⁶Cl uptake by basolateral and LMV, measured in the presence and absence of 2 mM DIDS. pH_i was varied from 6.1 to 8.0 by equilibrating the vesicles in media of appropriate HCO₃ concentration gassed with 95% N₂-5% CO₂ while keeping the extravesicular pH constant during the transport assays. In the BLMV (Fig. 4A), ³⁶Cl uptake displayed a sigmoidal pattern of activation as pH_i increased from 6.1 to 8.0, consistent with the presence of an internal H⁺ and/or OH modifier site on this exchanger. In contrast (Fig. 4B), uptake of ³⁶Cl by the luminal exchanger exhibited a concave velocity curve as pH_i increased from 6.1 to 8. At the acidic pH_i (6.7) prevailing in MTAL cells (14), the DIDS-sensitive component of ³⁶Cl uptake expressed relative to maximal uptake measured at pH_i 8.0 was threefold higher in LMV than in BLMV (30% vs. 10%).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Comparison of the effect of pHi on Cl/HCO3 exchange in BLMV (A) and LMV (B). Both types of vesicles were preincubated for 2 h at the indicated pH values (6.1-8.0) in media containing (in mM) 100 mannitol, 3 EGTA, and corresponding TMA HCO3 concentrations (1.1-87 mM). Osmolarity was maintained constant at 520 mosmol/kgH2O by using TMA gluconate. Various pHi values were obtained by using either 100 mM Tris-MES at pH 6.1 and 6.6 or with 100 mM Tris-HEPES at pH values of 7.0-8.0. Membrane vesicles were then rapidly diluted 1:11 and incubated for 9 s in a medium consisting of (in mM) 2 36Cl, 100 mannitol, 3 EGTA, 155 TMA gluconate, 100 Tris-MES (pH 5.5), and either 4 sodium gluconate () or 2 DIDS (); , difference between total and DIDS-insensitive 36Cl uptake. Both types of vesicles and reaction media were gassed with 95% N2-5% CO2. Values are means ± SE of 6 determinations on 3 separate preparations of BLMV and LMV.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Comparison of the effect of pHi on Cl/Cl exchange in BLMV (A) and LMV (B). Both types of vesicles were preincubated for 2 h at the indicated pH values (6.1-8.0) in media containing (in mM) 100 mannitol, 3 EGTA, 133 TMA gluconate, and 53 TMA Cl. Various pHi values were obtained as indicated in the legend to Fig. 4. The membrane vesicles were then rapidly diluted 1:31 and incubated for 9 s in a medium consisting of (in mM) 100 mannitol, 3 EGTA, 186 TMA gluconate, 100 Tris-MES (pH 5.5), and either 4 sodium gluconate () or 2 DIDS (). The final concentration of 36Cl was 3.2 mM; , difference between total and DIDS-insensitive 36Cl uptake. Both types of vesicles and reaction media were gassed with 100% N2. Values are means ± SE of 6 determinations on 3 separate preparations of BLMV and LMV.

Cl/HCO₃ antiporter kinetics in BLMV and LMV. Figure 6 compares the effect of external Cl concentration ([Cl]_o) on pH and HCO₃ gradient-stimulated ³⁶Cl uptake by BLMV and LMV measured in the presence and absence of 2 mM DIDS. In the presence of DIDS, best fits were obtained with a linear function (r = 0.99) for both BLMV (Fig. 6A) and LMV (Fig. 6B). When uptakes observed in the presence of DIDS were subtracted from total uptakes, curves were obtained in BLMV and LMV that described saturable Cl-dependent transport processes with Michaelis-Menten kinetics. Eadie-Scatchard plots of these data were consistent with the participation of a single Cl/HCO₃ antiport system in BLMV with an apparent K_m of 3.0 ± 0.5 mM and a V_max of 9.2 ± 2.2 nmol · mg¹ · 9 s¹ (Fig. 6A, inset). The corresponding values in LMV were an apparent K_m of 4.63 ± 0.61 mM and a V_max of 4.9 ± 0.6 nmol · mg¹ · 9 s¹ (Fig. 6B, inset). These values of apparent K_m and V_max did not significantly differ between BLMV and LMV, as determined by unpaired two-tailed t-tests.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Cl/HCO3 antiporter kinetics in BLMV (A) and LMV (B). Effect of increasing extravesicular concentrations of Cl (from 1 to 10 mM) on pH gradient- and HCO3 gradient-stimulated 9 s 36Cl uptake in BLMV and LMV in presence or absence of 2 mM DIDS (see Fig. 3 for composition of media). Total 36Cl uptakes measured in absence () and presence of DIDS () were used to calculate DIDS-sensitive uptake (). TMA salt concentration was maintained at 155 mM using appropriate concentrations of chloride and gluconate. Values are means ± SE of 6 determinations on 3 separate preparations of BLMV and LMV. Insets: Eadie-Scatchard plots of DIDS-sensitive 36Cl uptake in BLMV (A) and LMV (B).

View larger version (34K): [in this window] [in a new window]   Fig. 7.   Immunoblot analysis of AE2 and AE1 in two independent preparations of BLMV (B) and LMV (L). Each lane is loaded with 50 µg membrane protein. A: polyclonal rabbit antibody to mouse AE2 amino acids 1224-1237 detects predominant 165-kDa and less abundant 145-kDa AE2 polypeptides in BLMV but not in LMV. This antibody also detects additional polypeptides of lower Mr in BLMV and, at lower levels, in LMV [lane 1; longer exposure (not shown) confirms presence of these bands in lane 3, also]. All bands detected in the presence of irrelevant peptide (lanes 1-4) are competed by peptide antigen (lanes 5 and 6). B: monoclonal mouse antibody to rat AE1 fails to detect AE2 bands but detects all bands of lower Mr, confirming their identity as AE1 related. C: polyclonal antibody to mouse AE1 amino acids 917-929 fails to detect AE2 bands but detects all AE1-related bands of lower Mr. Asterisk in C indicates nonspecific reactivities not competed by peptide antigen (compare lanes 1-4 with lanes 5 and 6). Asterisk in B indicates band comigrating with this nonspecific band. AE2 and AE1 polypeptides were detected in 5 of 5 preparations examined. D: Na+-K+-ATPase -subunit is appropriately enriched in BLMV (the signal is oversaturated) but is present in nonnegligible amount in LMV, consistent with previously reported enzymatic assays (4). E: 31-kDa subunit of vacuolar H+-ATPase is appropriately enriched in LMV but also present in nonnegligible amount in BLMV.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

The objective of this study was to detect and characterize the Cl/HCO₃ exchangers on rat MTAL of Henle. To accomplish this, a new fractionation procedure was developed for the simultaneous isolation of LMV and BLMV from purified suspensions of rat MTAL, thereby permitting direct and separate characterization of the Cl/HCO₃ exchangers present at the luminal and basolateral cell surfaces of this nephron segment. The present study revealed that the rat MTAL possesses both luminal and basolateral Na⁺-independent Cl/HCO₃ exchangers (Fig. 1). Unlike the LMV exchanger, Cl/HCO₃ exchange by the BLMV transporter was markedly accelerated by increasing pH_i from 6.7 to 7.8, consistent with the presence of a proton-sensitive modifier site on this exchanger (Fig. 4). Both exchangers were DIDS-inhibitable; however, the BLMV exchanger was approximately fivefold more sensitive to DIDS (Fig. 3). The present study also established that a ~165-kDa AE2 polypeptide was detected only in BLMV, whereas AE1-related polypeptides were detected predominantly in BLMV and in lower abundance in LMV (Fig. 7). Our studies demonstrate for the first time the polarized distribution of distinct isoforms of anion exchanger polypeptides in the cells of the MTAL of Henle.

Earlier studies have failed to detect significant Cl/HCO₃ exchange in the absence of AVP and under isotonic conditions in mouse (23) and rat (24) MTAL suspensions by measuring pH_i. Leviel et al. (24) and Kikeri et al. (23) concluded that Cl/HCO₃ exchange was absent, because pH_i recovery in tubules incubated in the presence of an outwardly directed CO₂/HCO₃ gradient was insensitive to removal of external Cl. In both studies, however, substitution of chloride with gluconate was incomplete, resulting in final [Cl]_o levels ranging from 2.5 to 7.5 mM. These values of [Cl]_o likely supported significant Cl/HCO₃ exchange activity in MTAL, since in the current study the apparent K_m for [Cl]_o of the BLMV and LMV Cl/HCO₃ exchangers averaged 3 and 4.6 mM, respectively (Fig. 6). Previous studies of mouse AE2 in Xenopus oocytes have yielded low apparent K_m values of AE2 for [Cl]_o in Cl/Cl exchange mode of 5.6 mM, as determined from influx measurements (18), and 3.7 mM from efflux measurements (Y. Zhang and S. L. Alper, unpublished observations). Moreover, the Na⁺-containing solutions used in the experiments of Leviel et al. (24) and Kikeri et al. (23) may have enhanced alkali-loading processes (e.g., the BLMV and LMV Na⁺/H⁺ exchangers), thus counteracting acid-loading processes such as the Cl/HCO₃ exchangers of BLMV and LMV. These possibilities are supported by the subsequent preliminary studies of Sun and Dworkin (33), who detected basolateral DIDS-sensitive, Na⁺-independent and Cl-dependent HCO₃ transport in mouse MTAL segments perfused in vitro in the absence of AVP and under isotonic conditions. The possible presence of luminal Cl/HCO₃ exchange was not reported by these authors.

The major functional difference between the basolateral and apical anion exchanger activities was observed in the effects of pH_i (Fig. 4). The steep activation of BLMV ³⁶Cl/Cl exchange by increasing pH_i in nominally CO₂-free media (Fig. 5) suggested that internal H⁺ exerts a modifier effect on the BLMV Cl/HCO₃ exchanger. The presence of a strong internal modifier site for Cl/HCO₃ exchange reflected in steep pH_i dependence has previously been observed in rabbit parietal cell BLMV (27). A weaker internal modifier site, reflected in a broader pH_i dependence and decreased magnitude of activation, has also been noted in rabbit ileal brush-border membrane vesicles (26, 27). In contrast, ³⁶Cl/Cl exchange in rat MTAL LMV was unresponsive to pH_i values between 6.7 and 7.8 (Fig. 5). Thus the apparent pH_i dependence of Cl/HCO₃exchange in LMV (Fig. 4) presumably reflects its dependence on the intravesicular HCO₃ concentrations that varied in parallel with pH_i, as opposed to a modifier effect of OH/H⁺. Our observations that anion exchange in the BLMV isolated from rat MTAL displays a pH_i sensitivity consistent with regulation by an internal pH modifier site, whereas the apical AE is insensitive to pH_i, along with the uniquely basolateral distribution of AE2 (Fig. 7 and Ref. 2), are consistent with the known steep pH_i dependence of AE2 (18, 36). The latter studies found that AE2 and AE1 function in Xenopus oocytes differs markedly in their regulation by pH, with AE2 having a steeper pH dependence than AE1. However, unlike in the present study, combined variation of pH_i and pH_o were used, such that the "allosteric" control of AE2 may have been exerted by both pH_i and pH_o (36).

The polarized expression of different isoforms of the Cl/HCO₃ exchanger gene family in a nephron segment that resorbs NaCl and HCO₃ is consistent with the possibility that each isoform exerts distinct roles in the regulation of pH_i and cell volume as well as in vectorial ion transport. Recent studies found that the rat MTAL is the nephron segment with the highest concentration of AE2 transcript (6) and protein (2). Our current membrane fractionation studies (Fig. 7A) agree with previous immunocytochemical studies (2) that AE2 is located exclusively on the basolateral membrane of the rat MTAL, where the Na⁺/H⁺ exchanger NHE-1 is also expressed (4). This NHE isoform has been shown, in stable transfection studies, to be stimulated by hypertonicity (22), suggesting that AE2 and NHE-1 are the most likely candidates for cell volume regulation in the rat MTAL. Stimulation of NHE-1 by hypertonicity would alkalinize the cell and thereby activate the BLMV Cl/HCO₃ exchanger. Such a mechanism has already been proposed by Mason et al. (25) in lymphocytes and has also been described in Xenopus oocytes expressing heterologous AE2 (19). As was recently shown in oocytes (21), coupled operation of NHE-1 and AE2 produces a net NaCl gain that, together with osmotically obliged water, restores cell volume toward it original value.

In addition to its possible role in volume regulation, the presence of a pH-sensitive internal modifier site for BLMV Cl/HCO₃ exchange suggests that AE2 could have other functions in the rat MTAL. Although as assayed in the current experiments, the BLMV AE2 exchanger is maximally active only at alkaline pH_i values that are not reached physiologically in the cells of this nephron segment, it might in vivo undergo additional forms of regulation. This possibility is supported by recent evidence demonstrating that, at acidic pH_i values, NH⁺₄ at concentrations physiologically achieved in the MTAL activated AE2 expressed in Xenopus oocytes (17). Whether NH⁺₄ can also overcome the inhibitory effect of intracellular H⁺ similarly to activate AE2 of the rat MTAL at acidic pH_i remains to be determined.

At least two studies using isolated, perfused rat MTAL are consistent with the possibility that the BLMV anion exchanger of this nephron segment could play a role in HCO₃ transport and/or pH_i regulation. Watts and Good (35) found a background acid loading process, whose activity resembled that of the BLMV anion exchanger, with negligible activity at pH_i values below 6.7, which markedly increased to a maximal value at pH_i 7.5. Although these experiments were performed in the nominal absence of HCO₃, the pH_i-sensitive background acid loading process could have been due to Cl/HCO₃ exchange supported by low levels of ambient HCO₃ and/or to Cl/OH exchange. These processes presumably account for H⁺ gradient-stimulated Cl uptake in the nominal absence of HCO₃ (Fig. 1) and are potentiated in AE2 by hypertonicity (21).

More recently, Good et al. (14) found that 1 µM bath ethylisopropylamiloride (EIPA), a potent inhibitor of the Na⁺/H⁺ exchangers, inhibited HCO₃ absorption by 35%. They also found that 50 µM bath EIPA secondarily inhibited luminal Na⁺/H⁺ exchange activity, leading these authors to conclude for the existence of a functional coupling between basolateral and apical membrane NHEs. Importantly, these studies revealed that 1 µM bath EIPA significantly decreased pH_i from 7.10 to 7.05. Taken together, these findings raised the possibility that the coupling may be mediated by a change in cell volume as follows. Inhibition of the basolateral Na⁺/H⁺ exchanger would also inhibit the BLMV Cl/HCO₃ exchanger which, at pH_i of ~7.1, is exquisitely sensitive to small pH_i variations (Fig. 4). This would in turn inhibit basolateral NaCl entry via the parallel operation of the basolateral membrane Na⁺/H⁺ and Cl/HCO₃ exchangers, leading to cell shrinkage that secondarily could inhibit the luminal membrane Na⁺/H⁺ exchanger (35).

AE1-related polypeptides were also found in BLMV, in agreement with the preliminary report of the AE1 mRNA in rat MTAL (31), but in contrast to immunocytochemical studies (2). Since the pH_i dependence of known isoforms of AE1 (20, 36) is not consistent with that displayed by BLMV anion exchange (Figs. 4 and 5), the relationship between basolateral AE1 and H⁺-inhibited BLMV anion exchange remains unclear. It is possible, however, that the small DIDS-sensitive component of Cl uptake (0.80 ± 0.093 nmol · mg¹ · 9 s¹) observed in BLMV at pH_i 6.1 (Fig. 5A) represented operation of the basolateral AE1, since AE2-mediated ³⁶Cl uptake into Xenopus oocytes is absent at or below pH_o 6.2 (18), corresponding to a pH_i value of ~7.3 (36).

Although luminal AE1 has been proposed to mediate bicarbonate secretion by type B intercalated cells of collecting duct (1), the identity of this apical anion exchanger remains controversial (2, 11, 30). Similarly, interpretation of putatively apical AE1-related polypeptides in LMV of MTAL is complicated by several issues. First of these is the purity of the LMV fraction, since LMV are largely (but not completely) purified away from BLMV (Fig. 7D; and Ref. 4). Other possible sources of contamination in both LMV and BLMV fractions include eAE1 from residual red blood cell membrane and kAE1 from both plasmalemma and intracellular membranes of type A intercalated cells of medullary collecting duct.¹ Another problem is the comparison of the AE1-related polypeptides in LMV to the more abundant AE1-related polypeptides in BLMV, whose immunoblot signal is likely attenuated by very abundant, comigrating, -subunit of Na⁺-K⁺-ATPase (Fig. 7, C and D).

Nonetheless, Fig. 7 is consistent with the presence of AE1-related Cl/HCO₃ exchanger (AE1) in rat MTAL luminal membrane, where two NHE isoforms, NHE-2 (8, 34) and NHE-3 (3, 4), are also present. Since AE1 can operate at the acidic pH_i of MTAL cells, the Na⁺/H⁺ exchangers and the Cl/HCO₃ exchanger may "counteract" one another under physiological circumstances, providing negative feedback regulation of HCO₃ absorption. AE1 might also contribute to NaCl reabsorption via functional coupling with either NHE-2 and/or NHE-3, allowing to the MTAL cell the option to regulate independently NaCl absorption via coupled ion exchangers and NH⁺₄ absorption via the Na⁺-K⁺ (NH⁺₄)-2Cl cotransporter. Of note, in the isolated, perfused mouse (12) cortical thick ascending limb, but not in mouse MTAL (16), it has been shown that 50% of luminal NaCl absorption requires the presence of HCO₃. This HCO₃-dependent NaCl absorption is abolished by inhibition of carbonic anhydrase and inhibited by luminal SITS, suggesting that it is due to parallel action of Na⁺/H⁺ and Cl/HCO₃ exchange. Further studies are needed to examine the physiological import of these exchange modes in the apical membrane of the rat MTAL, a nephron segment containing high levels of carbonic anhydrase (10).

In conclusion, the present study directly demonstrates the presence of distinct Cl/HCO₃ exchangers in basolateral and apical plasma membranes isolated from rat MTALs. AE2 polypeptide is detected only in BLMV, whereas AE1-related polypeptides are detected predominantly in BLMV, and in lower, variable abundance in LMV. The most prominent functional difference between the anion exchangers in the two membrane fractions is the effect of pH_i. In contrast to the rather pH_i-insensitive LMV exchanger, the H⁺ concentration dependence of the BLMV exchanger suggests the existence of an intracellular H⁺ (OH) modifier site. Proton sensitivity of the BLMV Cl/HCO₃ antiport may play a role in pH_i homeostasis and volume regulation. We further propose that Cl/HCO₃ exchange in the MTAL basolateral membrane may contribute to HCO₃ absorption, whereas the luminal membrane exchanger may play a role in NaCl absorption via functional coupling with the luminal Na⁺/H⁺antiporters and/or may provide negative feedback regulation of HCO₃ absorption.