INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE TERMINAL inner medullary collecting duct (tIMCD) is the final nephron site of urinary acidification within the mammalian kidney. Ammonium (NH⁺₄) secretion in the tIMCD occurs, in part, through active proton transport in parallel with the passive diffusion of NH₃ (15). However, our laboratory has demonstrated an important role of direct NH⁺₄ transport in this segment. In cultured IMCD cells, NH⁺₄ and K⁺ compete for a common binding site on the Na⁺-K⁺-adenosinetriphosphatase (Na⁺-K⁺-ATPase) (34, 37). In native IMCD cells in suspension, both ions support ouabain-sensitive ATP hydrolysis (34). Thus both NH⁺₄ and K⁺ are transported directly by the Na⁺ pump.

To test the significance of Na⁺ pump-mediated NH⁺₄ uptake on proton secretion, the effect of ouabain on transepithelial net acid secretion was examined in rat tIMCD tubules perfused in vitro. Since deoxycorticosterone pivalate (DOCP) increases Na⁺ pump activity in the rat tIMCD (30), DOCP-treated rats were studied. In the absence of NH₄Cl, total CO₂ absorption (J_tCO2) was low and not affected by ouabain addition to the bath (30). Baseline J_tCO2 was higher in the presence of NH⁺₄ than in its absence (30). Moreover, in the presence of NH₄Cl, J_tCO2 was significantly inhibited upon ouabain addition to the bath (30). It was reasoned that NH⁺₄ enters the tIMCD cell on the basolateral membrane through an Na⁺-K⁺-ATPase-dependent pathway with release of protons in the cytosol. NH⁺₄ serves as a source of NH₃ and H⁺, which are secreted across the apical membrane. If such a model were true, then blockade of NH⁺₄ uptake through Na⁺ pump inhibition should decrease net proton entry and alkalinize the cell. To test this hypothesis, the effect of ouabain on intracellular pH (pH_i) was examined. Resting pH_i was lower in the presence than in the absence of NH₄Cl (30). Moreover, in the presence of NH₄Cl, addition of ouabain to the bath alkalinized the cell (30). Ouabain-induced alkalinization was not observed in the absence of NH₄Cl or when NH⁺₄ was substituted with another weak base (30). Thus NH⁺₄ and the Na⁺-K⁺-ATPase are important determinants of both net acid secretion and resting pH_i.

However, the mechanism for the increase in pH_i and the reduction in J_tCO2 observed following ouabain addition could be due to blockade of Na⁺-K⁺-ATPase-mediated NH⁺₄ uptake or through changes in ion gradients generated by the Na⁺ pump. The Na⁺ pump could generate ion gradients that affect other NH⁺₄/OH/H⁺/ transporters independent of direct Na⁺-K⁺-ATPase-mediated NH⁺₄ uptake. The purpose of this study was to further characterize the NH⁺₄-dependent increase in pH_i and reduction in J_tCO2 that follow ouabain addition and to determine whether these observations can be explained by a transport mechanism other than Na⁺-K⁺-ATPase-mediated NH⁺₄ uptake.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Tissue preparation. tIMCD tubules were dissected from pathogen-free male Sprague-Dawley rats weighing 65-120 g (Rm. 205G; Harlan, Indianapolis, IN). All animals were housed in microisolator cages and fed a low-Na⁺, 0.8% K⁺ diet (Ziegler Brothers, Gardners, PA) (36). Rats were injected with 5 mg DOCP (CIBA-Geigy Animal Health, Greensboro, NC) by intramuscular injection 5-7 days prior to death. DOCP was employed to increase Na⁺-K⁺-ATPase activity, as described previously (30). Animals were injected with furosemide (5 mg/100 g body wt ip) 45 min before death by decapitation to induce a rapid diuresis. This furosemide-induced diuresis reduces the inner medullary axial solute concentration gradient (36) and attenuates changes in the extracellular osmolality of the tubule.

Unless otherwise stated, all experiments were performed in bicarbonate-buffered solutions (Table 1), gassed with 95% air-5% CO₂ before use. The measured osmolalities of all solutions are listed (Table 1).

Coronal slices were cut from the kidneys and placed into a dissection dish containing the chilled experimental solution (11°C). IMCDs were dissected from the middle third of the inner medulla as described previously (36). Tubules were mounted on concentric glass pipettes and perfused in vitro at 37°C. Experiments were performed with identical solutions in the perfusate and bath. In some experiments, ouabain (2.5 or 5 mM) or bumetanide (100 µM) was added to the bath fluid only. To maintain the desired CO₂ concentration, the perfusate was passed through jacketed concentric tubing through which 95% air-5% CO₂ was blown in a countercurrent direction around the perfusate line. To maintain pH in -containing solutions, the bath fluid was constantly bubbled with 95% air-5% CO₂. In N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES)-buffered solutions, bath fluid was bubbled with 100% O₂. Bath pH was measured continuously during all experiments as described previously (36). Bumetanide was prepared as a 100 mM stock in 500 mM tris(hydroxymethyl)aminomethane (Tris). Ouabain was dissolved directly into the bath solution.

Measurement of bicarbonate flux. Tubule fluid samples were collected under oil in calibrated constriction pipettes. Flow rate was determined as described previously (36). Total CO₂ (tCO₂) concentration was measured in the collected fluid (C_L) and perfusate (C_o) using a continuous flow fluorometer (30). The CO₂ reagent was purchased as a kit (no. 132-A; Sigma, St. Louis, MO) and diluted to 50% strength with water. Using this method, bicarbonate (total CO₂, tCO₂) concentration differences of less than 1 mM can be detected using a pipette of 8 nl (30). Bicarbonate flux, J_tCO2, was calculated according to the equation where C_o and C_L are the perfusate and collected fluid tCO₂ concentration. V_L is the flow rate (in nl/min), and L is the tubule length (in mm). This equation assumes zero net fluid transport in the absence of an imposed osmolality gradient (36). To eliminate detection of tCO₂ flux which represents artifact generated by loss of CO₂ in the collection pipette, C_o was estimated by measuring tCO₂ concentration in the collected fluid at very fast flow rates (>50 nl · mm¹ · min¹). Thus CO₂ loss was matched in C_o and C_L measurements, allowing the loss terms to cancel (36). Amiloride at a 1 µM concentration did not affect the fluorescence signal of the CO₂ reagent (n = 3, data not shown).

Measurement of pH_i. pH_i was measured using the esterified form of 2',7'-bis(carboxyethyl)-5(6)-carboxyfluorescein (BCECF) (30). The detailed methodology for measurement of pH_i in tubules perfused in vitro has been described previously (30).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Effect of limiting Na⁺ entry on J_tCO2 and resting pH_i. It was hypothesized that across the basolateral membrane of the rat tIMCD, NH⁺₄ uptake is mediated by the Na⁺-K⁺-ATPase. NH⁺₄ thus provides a source of H⁺ and NH₃ for luminal secretion and the titration of other luminal buffers (30). In the presence of NH₄Cl, the reduction in J_tCO2 and the increase in pH_i observed following ouabain addition (30) can be explained by inhibition of Na⁺ pump-mediated NH⁺₄ uptake with reduced net H⁺ entry and increased pH_i. However, Na⁺ pump inhibition affects other H⁺/OH//NH⁺₄ pathways. Ouabain-induced changes in activity of these other transporters could also explain the above observation. The purpose of this study was to explore further the mechanism responsible for the NH⁺₄-dependent, ouabain-induced reduction in J_tCO2 and increase in pH_i observed previously (30).

1

1

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Effect of luminal amiloride on total CO2 absorption (JtCO2) with NH4Cl present in bath and perfusate. A: effect of 1 µM amiloride on JtCO2. During control period, mean flux was 3.7 ± 0.7 pmol · mm1 · min1. After addition of 1 µM amiloride to the luminal perfusate, flux declined to 2.0 ± 0.9 (n = 5, P < 0.05). B: experiment in A was repeated, but with a mock perfusate exchange, i.e., amiloride was not introduced into luminal perfusate. Baseline JtCO2 was 3.4 ± 1.1 and 3.6 ± 0.8 pmol · mm1 · min1 after perfusate exchange (n = 3, P = NS).

Effect of limiting Na⁺ entry on NH⁺₄-dependent, ouabain-induced changes in pH_i. If Na⁺ pump-mediated NH⁺₄ entry supplies the cell cytosol with H⁺ and NH⁺₄, then Na⁺ pump blockade should limit NH⁺₄ uptake, attenuating net H⁺ entry and alkalinizing the cell. The effect of ouabain on resting pH_i was therefore explored. Results were compared when the experiment was repeated with substitution of ouabain for its vehicle (solution 1). In the presence of 3 mM KCl, 6 mM NH₄Cl, and 25 mM NaHCO₃/5% CO₂ in the bath and perfusate (solution 1), resting pH_i averaged 7.18 ± 0.02 (n = 16). As shown in Fig. 2 (solution 1), with ouabain addition to the bath, a prompt and sustained increase in pH_i was observed. Since pH_i remained elevated, compared with control (vehicle), for at least 5.5 min following the addition of ouabain to the bath, pH_i was measured 4 min after the addition of ouabain to the peritubular bath. These results confirm our previous observations that ouabain addition to the bath results in increased pH_i when perfused in the presence of NH₄Cl (30). The effect of ion substitution on ouabain-induced alkalinization was studied when each tubule was used as its own control.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Effect of ouabain on resting intracellular pH (pHi) with NH4Cl present in bath and perfusate. Tubules were perfused with symmetrical solutions containing 3 mM KCl + 6 mM NH4Cl (solution 1). Bath was exchanged with the introduction of 2.5 mM ouabain and pHi measured every 45 s. Results are the mean of 3 tubules studied. In separate tubules, the experiment was repeated, but with a mock bath exchange and without introduction of ouabain. Ouabain vehicle was solution 1. Results are the mean of 3 experiments.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Effect of limiting Na+ entry on NH+4-dependent, ouabain-induced alkalinization. Tubules were perfused and bathed with NH4Cl present in bath and perfusate (solution 1). Baseline pHi was measured. Bath solution was exchanged to include ouabain (2.5 mM), and pHi was measured 4 min after the exchange. Bath was again exchanged, removing ouabain, and pHi was measured 4 min later (recovery). In the same tubules, perfusate and bath was exchanged, removing extracellular Na+ (solution 2). Ouabain-induced alkalinization was measured as above. Dotted line indicates that the experiment was performed in the reversed order, i.e., ouabain-induced alkalinization was measured first in absence and then in presence of extracellular Na+.

Effect of extracellular K⁺ on ouabain-induced alkalinization. Our laboratory has demonstrated that NH⁺₄ and K⁺ are competitive inhibitors for a common extracellular binding site of the Na⁺-K⁺-ATPase (34, 37). Thus increased extracellular K⁺ concentration attenuates NH⁺₄ uptake by the tIMCD cell (34, 37). Using the model of Kurtz and Balaban (18) and employing kinetic values measured in our laboratory (34), we were able to estimate NH⁺₄ uptake through the Na⁺-K⁺-ATPase. The model predicts NH⁺₄ uptake through the Na⁺ pump to be increased two- to threefold when the extracellular K⁺ concentration is reduced from 12 to 3 mM. We reasoned that if the model were true, then in the presence of NH₄Cl, increasing extracellular K⁺ concentration should attenuate ouabain-induced alkalinization as well as total and ouabain-sensitive J_tCO2. To test this hypothesis, NH⁺₄-dependent, ouabain-induced alkalinization was measured at two extracellular K⁺ concentrations (3 and 12 mM). In the presence of NH₄Cl, pH_i was 7.10 ± 0.06 at a K⁺ concentration of 3 mM (solution 1, Table 4) and 7.16 ± 0.05 (n = 4) at a K⁺ concentration of 12 mM (solution 4, Table 4). These results compare with J_tCO2 flux rates of 3.4 ± 0.4 (n = 11) and 2.5 ± 0.3 pmol · mm¹ · min¹ (n = 7) when the extracellular K⁺ concentrations were 3 and 12 mM, respectively (Table 2). Thus, on average, J_tCO2 was increased, and resting pH_i was reduced, at lower extracellular K⁺ concentrations. These differences, however, did not reach statistical significance.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Effect of increased extracellular K+ concentration on ouabain-induced alkalinization. Ouabain-induced alkalinization was measured in presence of 3 mM KCl + 6 mM NH4Cl (solution 1), as outlined in Fig. 3. Ouabain concentration was 5 mM. Perfusate and bath were then exchanged to solution containing 12 mM KCl + 6 mM NH4Cl (solution 4), and ouabain-induced alkalinization was again measured. In 2 tubules, the experiment was performed in the reverse order, i.e., ouabain-induced alkalinization was first measured with 12 mM KCl + 6 mM NH4Cl present. Results are means ± SE of 4 experiments.

1

1

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Effect of ouabain on JtCO2 when the extracellular K+ concentration is increased. A: effect of ouabain on JtCO2 in presence of 3 mM KCl + 6 mM NH4Cl. During control period, mean flux was 3.8 ± 0.5 pmol · mm1 · min1. After addition of 2.5 mM ouabain to the bath, flux declined to 1.6 ± 0.3 pmol · mm1 · min1 (n = 7, P < 0.05). These data were taken from Ref. 30. B: experiment in A was repeated, but in presence of 12 mM KCl + 6 mM NH4Cl (solution 4). Baseline JtCO2 was 2.9 ± 0.4 and 2.9 ± 0.3 pmol · mm1 · min1 following the addition of 5 mM ouabain to bath (n = 4, P = NS).

Role of Ba²⁺-sensitive K⁺ channels. Increasing extracellular K⁺ concentration could increase pH_i by limiting Na⁺ pump-mediated NH⁺₄ entry. An alternative hypothesis, however, is that increasing extracellular K⁺ depolarizes the cell, which alters the activity of other H⁺/OH transporters. If this hypothesis were true, then NH⁺₄-dependent, ouabain-induced alkalinization results from membrane depolarization, which in turn alters electrogenic H⁺/OH transport rather than Na⁺-K⁺-ATPase-mediated NH⁺₄ uptake. At a concentration of 1 mM, BaCl₂ induces a rapid and sustained cellular depolarization through K⁺ channel blockade in rat IMCD tubules perfused in vitro (26). Thus the effect of BaCl₂ on pH_i was explored to test the effect of changes in membrane potential on pH_i, independent of direct Na⁺ pump blockade or changes in extracellular K⁺ concentration. Resting pH_i was examined in the presence and the absence of Ba²⁺ (solution 5). Following the addition of 1 mM Ba²⁺ to the peritubular bath (n = 5), pH_i was similar to that observed in time controls (n = 3, solution 5, Ba²⁺ absent; Fig. 6 a nd Table 5). Thus Ba²⁺ does not induce cellular alkalinization commensurate with that observed upon ouabain addition (solution 5; Fig. 6B and Table 5).² Moreover, NH⁺₄-dependent, ouabain-induced alkalinization was not abolished when 1 mM BaCl₂ was present in the bath solution (Table 5).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Effect of Ba2+ on resting pHi. A: tubules were perfused in presence of 3 mM KCl + 6 mM NH4Cl but in absence of and (solution 5) (see footnote 2.). Then, baseline pHi was measured (left). Bath solution was exchanged with the introduction of 1 mM BaCl2. pHi was measured 4 min later. In a time control series (right), the experiment above was repeated, but NaCl was substituted for BaCl2 isosmotically (n = 3). B: ouabain-induced alkalinization was measured under the same conditions as above (solution 5), but in absence of Ba2+. Baseline pHi was measured. pHi was again measured 4 min after ouabain addition (2.5 mM) and then 4 min following ouabain withdrawal.

Role of transport. Another explanation for the observed NH⁺₄-dependent reduction in J_tCO2 and increase in pH_i observed following ouabain addition is through changes in transport. The Cl/ exchanger and Na⁺- symporter are located on the basolateral membrane of the rat tIMCD (12, 33). In other cell types such as the proximal tubule, NH₃ uptake "back-titrates" intracellular protons and thus increases pH_i (29). The NH₃-induced increase in pH_i could stimulate exit through either of the above pathways. Moreover, in other cells the Na⁺ pump modulates both Cl/ exchange and Na⁺- symport activity (27, 31). Thus the increase in pH_i and the reduction in J_tCO2, which follow ouabain addition, could be explained by changes in either Na⁺- or Cl/-mediated efflux.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Effect of ouabain on resting pHi in absence of /CO2. Tubules were perfused and bathed in presence of 3 mM KCl + 6 mM NH4Cl but in absence of /CO2 (solution 6). Baseline pHi was measured. Resting pHi was then measured following addition and then withdrawal of ouabain (2.5 mM) to peritubular bath. Bath was then exchanged to include 0.1 mM ethoxzolamide, a carbonic anhydrase inhibitor. Resting pHi was again measured in presence and absence of ouabain (2.5 mM) in peritubular bath. Results from 5 separate tubules are displayed.

View larger version (10K): [in this window] [in a new window]   Fig. 8.   Effect of ouabain on resting pHi in absence of both /CO2 and NH4Cl. Tubules were perfused and bathed in presence of 3 mM KCl, but without NH4Cl (solution 7). Baseline pHi was measured. pHi was again measured following addition and then withdrawal of ouabain (2.5 mM) in peritubular bath. Results obtained from 3 separate tubules are displayed.

Effect of Na⁺-K⁺-2Cl inhibition on J_tCO2 and resting pH_i. Our laboratory has shown that NH⁺₄ and K⁺ compete for a common binding site on the Na⁺-K⁺-2Cl cotransporter in mouse tIMCD cells (37). Furthermore, both ions are transported through this carrier (37). These results are in keeping with a recent report of high levels of Na⁺-K⁺-2Cl cotransport expression in the mouse tIMCD (14). In many cell types, increased Na⁺_i reduces K⁺ (or NH⁺₄) uptake through the Na⁺-K⁺-2Cl cotransporter (27, 40). Thus the reduction in J_tCO2 and the increase in resting pH_i observed with ouabain addition could occur from reduced Na⁺-K⁺-2Cl cotransport-mediated NH⁺₄ uptake. If this hypothesis were true, then addition of an Na⁺-K⁺-2Cl cotransport inhibitor, such as bumetanide, should inhibit J_tCO2 the same or more than that observed with ouabain addition. Bumetanide at a concentration of 100 µM fully inhibits the Na⁺-K⁺-2Cl cotransporter in the rat tIMCD (11). Thus bumetanide in a 100 µM concentration was employed, and its effects on J_tCO2 and resting pH_i were tested (Table 2; Fig. 9). Our laboratory reported previously that in tIMCD tubules from DOCP-treated rats perfused in vitro in the presence of 3 mM KCl + 6 mM NH₄Cl (solution 1), J_tCO2 fell from 3.8 ± 0.5 to 1.6 ± 0.3 pmol · mm¹ · min¹ upon the addition of ouabain to the bath (n = 7, P < 0.05) (30). We asked ether the reduction in J_tCO2 observed following ouabain addition could be reproduced with bumetanide. Under these conditions (solution 1), baseline J_tCO2 was 3.1 ± 0.4 and 2.8 ± 0.5 pmol · mm¹ · min¹ with the application of 100 µM bumetanide to the bath (n = 3, P = NS). Thus, although ouabain addition to the bath reduced J_tCO2, the addition of bumetanide did not.

View larger version (13K): [in this window] [in a new window]   Fig. 9.   Effect of bumetanide on JtCO2. Tubules were perfused and bathed in presence of 3 mM KCl + 6 mM NH4Cl (solution 1). During control period, mean flux was 3.1 ± 0.4 pmol · mm1 · min1. After addition of 100 µM bumetanide to peritubular bath, flux was 2.8 ± 0.5 pmol · mm1 · min1 (n = 3, P = NS).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Our laboratory has demonstrated previously that net acid secretion in the tIMCD is greater in the presence of NH₄Cl than in its absence (30). Furthermore, in the presence but not in the absence of NH₄Cl, ouabain addition reduces net acid secretion and increases pH_i. These observations cannot be explained by differences in buffering of the luminal fluid or the cytosol which might occur in the presence and absence of NH₄Cl³ (30). The present study demonstrates that the NH⁺₄- dependent, ouabain-induced alkalinization and reduction in J_tCO2 occur through inhibition of NH⁺₄ uptake by the Na⁺-K⁺-ATPase. We conclude that Na⁺-K⁺-ATPase-mediated NH⁺₄ uptake is an important determinant of pH_i and net acid secretion in the rat tIMCD.

In mouse tIMCD cells in culture, a K⁺/NH⁺₄ exchanger has been reported (1). Since the driving force for K⁺ exit/NH⁺₄ uptake, or K⁺/NH⁺₄ exchange (Fig. 10) (1), is decreased with Na⁺ pump inhibition, NH⁺₄- dependent, ouabain-induced alkalinization might be explained by changes in K⁺/NH⁺₄ exchange. This question cannot be tested directly, since no specific inhibitors of K⁺/NH⁺₄ exchange are available (1, 38). However, it is unlikely that ouabain-induced changes in pH_i and J_tCO2, as observed in this study, are mediated by changes in K⁺/NH⁺₄ exchange activity. Inhibition of the Na⁺ pump decreases the driving force for K⁺/NH⁺₄ exchange by reducing intracellular K⁺ (K⁺_i) concentration. In many cells, including the proximal tubule (22, 41), following the addition of ouabain, Na⁺_i increases, and hence K⁺_i decreases, over a time period of at least 20 min. However, as shown in Fig. 2, the alkalinization observed following ouabain addition is complete within 2 min. Thus pH_i changes that follow ouabain addition do not parallel expected changes in K⁺_i. Moreover, we have shown that both NH⁺₄ and K⁺ support equivalent rates of ouabain-sensitive ATP hydrolysis in permeabilized, native rat IMCD cells (34). Thus both cations are transported directly by the Na⁺-K⁺-ATPase. Since these experiments were performed in permeabilized cells, inhibition of NH⁺₄ transport following ouabain addition occurred independent of changes in intracellular ion composition.

In the rat tIMCD, two K⁺-ATPases have been identified (16). One of these K⁺-ATPases, like the gastric H⁺-K⁺-ATPase, is sensitive to low concentrations of Sch-28080 but insensitive to ouabain. The other K⁺-ATPase is insensitive to Sch-28080 but sensitive to ouabain and thus resembles the "colonic" H⁺-K⁺-ATPase isoform. NH⁺₄ transport by the H⁺-K⁺-ATPase has been described (8). Therefore, it is possible that the ouabain-sensitive "colonic" H⁺-K⁺-ATPase transports NH⁺₄ and therefore mediates the NH⁺₄-dependent, ouabain-induced increase in pH_i and decrease in J_tCO2 observed in the present study. However, activity of the colonic H⁺-K⁺-ATPase is insensitive to changes in Na⁺ availability (4) and therefore cannot explain the observations of the current study. We have reported Sch-28080-sensitive bicarbonate absorption in the rat tIMCD (38). The transporter mediating Sch-28080-sensitive bicarbonate absorption, however, is distinct from the transport mechanism that mediates the NH⁺₄-dependent, ouabain-sensitive changes in pH_i and J_tCO2 described herein (38).

Our results do not support significant K⁺ channel-mediated NH⁺₄ efflux on the basolateral membrane of the rat tIMCD. However, they cannot exclude channel-mediated NH⁺₄ efflux on the apical membrane. ATP-sensitive K⁺ channels sensitive to intracellular but not extracellular Ba²⁺ have been localized to the apical membrane of the mouse IMCD (24). Similarly, a nonspecific, amiloride-sensitive cation channel on the apical membrane has been reported (20). Both of these channels transport NH⁺₄. Thus depolarization of the cell with ouabain addition could facilitate NH⁺₄ efflux across the apical membrane mediated by these channels. The result would be increased NH⁺₄ secretion into the luminal fluid upon ouabain addition. However, in a previous study (30), when ouabain was applied to the peritubular bath, total ammonium secretion fell from 0.3 ± 0.1 to 0.1 ± 0.1 pmol · mm¹ · min¹. Thus ouabain does not increase channel-mediated NH⁺₄ efflux across the apical membrane. Thus changes in K⁺ channel-mediated NH⁺₄ transport cannot explain the observations of the present study.

In epithelia that utilize carbonic acid as the main proton source, net luminal proton secretion occurs when proton secretion across the apical membrane is accompanied by bicarbonate (base) secretion across the basolateral membrane (31). However, cytosolic carbonic anhydrase activity is less abundant in the tIMCD than in other nephron segments (32, 33). Therefore, the regulation of basolateral exit and apical proton secretion in parallel may be less important in the tIMCD than in other nephron segments. Our results show that NH⁺₄, rather than carbonic acid, provides the primary source of protons in the tIMCD.

View larger version (22K): [in this window] [in a new window]   Fig. 10.   Proposed NH+4 pathway in rat terminal inner medullary collecting duct (tIMCD). NH+4 is taken up on basolateral membrane by the Na+-K+-ATPase. Intracellular NH+4 provides a source of H+ and NH3 for the cell. H+ is secreted across the apical membrane through the H+-K+-ATPase. NH3 leaves the cell by passive diffusion across the apical membrane, or it recycles across the basolateral membrane. Other NH+4/OH/H+/ transport pathways reported in tIMCD are shown (1, 11-13, 20, 24, 26, 28, 30, 35, 38).

In the tIMCD, NH⁺₄ is an important H⁺ source. Our data show that NH⁺₄ is taken up across the basolateral membrane by the Na⁺ pump, providing a source of H⁺ and NH₃. Protons are then secreted across the apical membrane. However, since the pK_a for NH₃/NH⁺₄ is 9.03 (15), in the range of physiological pH_i, only ~1% of NH⁺₄ releases a proton (39). Thus NH⁺₄ uptake alone should not result in measurable changes in pH_i. However, rapid entry of NH⁺₄ coupled with NH₃ exit across the basolateral membrane should provide significant net proton uptake with a substantial fall in pH_i (30, 39). Across the basolateral membrane, NH⁺₄ uptake with NH₃ efflux would provide an NH₃ shuttle, with net proton uptake. Across the apical membrane, NH₃ secretion in parallel with H⁺ secretion by the H⁺-K⁺-ATPase (or H⁺-ATPase) (38) would trap NH⁺₄ and facilitate net acid secretion (Fig. 10).

Amiloride administration decreases K⁺ excretion and impairs urinary acidification (5, 7). These defects in H⁺ and K⁺ secretion generate a hyperkalemic, hyperchloremic metabolic acidosis (7). In the cortical collecting duct (21) and turtle bladder (2), amiloride administration reduces the lumen-negative potential difference and decreases proton secretion. The acidification defect that follows amiloride administration is felt to occur from elimination of this lumen-negative potential difference, which decreases the driving force for H⁺ secretion. In the turtle bladder, if the lumen-negative potential difference is restored, then the defect in proton secretion that follows amiloride administration is reversed, despite the ongoing presence of amiloride (2). Thus the amiloride-induced reduction in proton secretion has been attributed to a "voltage defect." In the rat tIMCD, micropuncture studies have demonstrated a reduction in net acid secretion with luminal amiloride (5). In bicarbonate-loaded rats, DuBose and Caflisch (5) observed that the papillary urine-to-blood PCO₂ gradient, an index of proton secretion, was reduced with amiloride administration. The present study demonstrates a reduction in J_tCO2 upon addition of amiloride to the perfusate that cannot be explained by changes in V_T. The present study, therefore, provides an explanation for this impaired proton secretion in the tIMCD. Amiloride reduces Na⁺_i availability, which attenuates Na⁺ pump-mediated NH⁺₄ uptake and therefore reduces net proton secretion.⁴ Luminal acidification in the tIMCD is thus dependent on apical Na⁺ entry.

Hyperkalemia is frequently associated with metabolic acidosis (6). Hyperkalemia reduces ammonium production by the proximal tubule and reduces ammonium absorption by the thick ascending limb, which together attenuate net acid excretion (6). In rats receiving dietary K⁺ loading, DuBose and Good (6) have observed a decrease in the interstitial NH₃ concentration and therefore a reduction in the NH₃ gradient from the interstitium to the collecting duct lumen. However, in rats which received a high-K⁺ diet, no net transfer of ammonium from the interstitium to the lumen was observed, despite the presence of an NH₃ gradient that should favor NH₃ diffusion into the collecting duct lumen. This observation suggested a defect in NH⁺₄ transfer from the interstitium and the collecting duct lumen. The present study shows that this transfer defect can be explained, at least in part, by competition between NH⁺₄ and K⁺ for the extracellular binding site on the Na⁺ pump. Increasing extracellular K⁺ reduces NH⁺₄ uptake by the tIMCD cell and hence reduces acid secretion.

In conclusion, NH⁺₄ uptake across the basolateral membrane of the tIMCD is mediated by the Na⁺ pump. NH⁺₄ uptake provides a source of H⁺ for apical H⁺ secretion and the titration of luminal buffers. The reduction in acid secretion in the rat tIMCD observed following the administration of amiloride or upon dietary K⁺ loading can be explained, at least in part, by reduced NH⁺₄ uptake by the Na⁺-K⁺-ATPase.