Extracellular K+-dependent H+ extrusion after an acute acid load, an index of H/K exchange, was monitored in intercalated cells (ICs) from rat cortical collecting tubule (CCT) using ratiometric fluorescence imaging of the intracellular pH (pHi) indicator, 2′,7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF). The hypothesis tested was that 12- to 14-day NaCl deprivation increases H-K-ATPase in rat ICs. The rate of H/K exchange in the low-NaCl ICs was double that of controls. In control ICs, H/K exchange was inhibited by Sch-28080 (10 μM). In the low-NaCl ICs, it was partially blocked by Sch-28080 or ouabain (1 mM). Simultaneous addition of both inhibitors abolished K-dependent pHirecovery. The induced H/K exchange observed with NaCl restriction was not due to elevated plasma aldosterone as evidenced by experiments on ICs from rats implanted with osmotic minipumps administering aldosterone such that plasma levels were comparable to those of NaCl-deficient rats. The results suggest that NaCl deficiency induces two isoforms of H-K-ATPase in ICs and that this effect is not mediated by elevated plasma aldosterone.
- proton/potassium exchange
it is well established that an H-K-ATPase, a member of the P-type ATPase gene family, exists in the mammalian kidney (8, 40). Functional, immunocytochemical, pharmacological, and biochemical evidence (40) suggests that one of the H-K-ATPase isoforms identified in the distal nephron is most closely related to the H-K-ATPase found in the stomach. This isoform is also inhibited by the imidazopyridine, 2-methyl-8-(phenylmethoxy)imidazo[1,2a]-pyridine-3-acetonitrile (Sch-28080), a specific inhibitor of the gastric H-K-ATPase (33). In the kidney, it is thought to play a role in both K reabsorption and acid secretion (11, 13, 26, 27, 39) and functions in collecting duct under basal conditions (27). In addition, a ouabain-sensitive isoform of H-K-ATPase appears to be expressed in the kidney with K depletion (3). mRNA levels of colonic H-K-ATPase have been measured in cortical collecting tubule (CCT) (31) and renal cortex (23, 35) of control rats and kidney from K-depleted rats (7, 14, 16). During K depletion (39) and chronic metabolic acidosis (26) the physiological significance of the differential expression of the ouabain-sensitive colonic isoform in the distal nephron has yet to be determined.
Hormonal regulation of renal H-K-ATPase function is also uncertain. Early reports suggested a role for aldosterone in stimulating H-K-ATPase activity (11). In addition, studies using the renal cell line MDCK demonstrated an omeprazole-sensitive H/K exchanger induced with chronic exposure to aldosterone (20). In contrast, Eiam-Ong et al. (9) saw no effect of elevated plasma aldosterone levels on H-K-ATPase activity in microdissected nephron segments. These investigators did observe a direct correlation between plasma K status and H-K-ATPase activity in the nephron segments. In addition, Jaisser et al. (14) did not see any changes in colonic K-ATPase mRNA levels in kidney with an elevation in plasma aldosterone.
The purpose of this investigation was to measure H/K exchange at the single cell level in split-open rat CCT using ratiometric digital imaging techniques. We tested the hypothesis that chronic NaCl deprivation increases H-K-ATPase activity in rat intercalated cells (ICs). Our results demonstrate a Sch-28080-sensitive H/K exchanger in ICs from control rats and rats maintained on a NaCl-deficient diet. In addition the ICs from the NaCl-restricted rats express a ouabain-sensitive H/K exchanger that is not apparent in ICs from control rats. Furthermore, ICs from the NaCl-restricted rats have an increased rate of H/K exchange compared with controls. This response does not appear to be mediated by changes in plasma acid-base or K status or elevated plasma aldosterone levels.
MATERIALS AND METHODS
Pathogen free Sprague-Dawley rats of either sex (Charles River Laboratories, Kingston, NY) weighing between 100–150 g were used for these experiments. Rats were fed either a normal diet (Purina Formulab 5008, Na content 2.8 g/kg, K content 11 g/kg) or a low-NaCl diet (diet 902902, Na content 3.8 mg/kg, K content 8.6 g/kg; ICN Biochemicals, Cleveland, OH) for 12–14 days. In one series of experiments, osmotic minipumps were implanted subcutaneously (model 2002; Alza, Palo Alto, CA). The pumps contained aldosterone (Sigma Chemical, St. Louis, MO) dissolved in polyethylene glycol 300, which were administered at an infusion rate of 250 μg ⋅ kg body wt−1 ⋅ day−1for 12–14 days; a plasma level expected is comparable to that measured in plasma of rats maintained on a low-NaCl diet (21).
Tubules were prepared and mounted in the chamber as previously described (27). Briefly, rats were killed by cervical dislocation, the kidneys were removed, and CCTs were dissected free and opened to form a flat epithelium. Sometimes two tubules were used from each animal.
Solutions were gravity fed into a manually operated six-port Hamilton valve. The solution leaving the valve went directly into a miniature water-jacketed glass coil (Radnotti Glass Technology, Monrovia, CA) for regulating the solution temperature. The warmed solution entered the experimental chamber, which was mounted on the stage of an inverted epifluorescence microscope (Nikon Diaphot). The temperature of the superfusate in the chamber was maintained at 37°C.
Tubules were superfused with HEPES-buffered solutions as described in Table 1. Bafilomycin A1 (LC Labs, New Bedford, MA; and Alexis, San Diego, CA) was dissolved in DMSO to 1 mM and diluted 1:10,000 for a final concentration of 100 nM in the experimental superfusates. All chemicals were obtained from Sigma Chemical unless otherwise stated. Nigericin (Molecular Probes, Eugene, OR) was added to potassium Ringer solutions (solutions a and b) from a 10 mM stock (3 parts ethanol, 1 part dimethylformamide) for a final concentration of 10 μM. Individual vials (50 μg) of the acetoxymethyl derivative of 2′,7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF-AM; Molecular Probes) were stored dry at 0°C and reconstituted in DMSO (at a concentration of 10 mM) for each experiment. The final loading concentration of dye was 5 μM in sodium Ringer solution. Sch-28080, a gift from Dr. A. Barnett at Schering Plough, was dissolved in methanol and added to the appropriate solutions to a final concentration of 10 μM to yield a 0.1% (volume to volume) concentration of vehicle. Ouabain (1 mM) was added directly to the superfusing solutions where indicated.
Equipment. The basic components of the experimental apparatus consisted of the following: an inverted epifluorescence microscope (Nikon Diaphot) equipped with a 75-W xenon lamp, Nikon CF Fluor ×100/1.3 NA, oil-immersion objective, fluorescence excitation filter wheel (Metal Tek) coupled to an additional filter wheel (Metal Tek) holding a variety of neutral density filters, computer controllable excitation light shutter, and a cooled, charge-coupled device (CCD) camera (Princeton Instruments) with a frame transfer chip (EEV-37) and 12-bit readout. In addition, for some experiments a CCD camera (CCD 200, Videoscope International) attached to a VS2525 Image Intensifier (Videoscope) was used. The emitted fluorescence signal is relayed as real time continuous output to an IBM PC/AT compatible clone, and the image pairs were collected on a Sierra Pinnacle Micro optical disc drive (1.3 gigabyte) where data can be stored for future analysis on a pixel-by-pixel basis using the Metafluor software package (Universal Imaging). Quantitative image pairs at 490- and 440-nm excitation, with emission at 520 nm, were obtained every 15 s over a period of about 30 min. Neutral density filters coupled to the excitation filters minimized the transmitted excitation light to between 1 and 10%, depending upon the experiment, to minimize photobleaching and photodynamic damage to the cells. The individual wavelength intensities were stable over the experimental time course. The fluorescence excitation was shuttered off, except during the brief periods required to record an image. To correct for intrinsic autofluorescence and background, images were obtained using the experimental acquisition configuration on the split-opened portion of tubule prior to loading with the dye. These background images were subtracted from the corresponding images of cell fluorescence, and background corrected ratios were generated for data analysis.
BCECF loading and identification of cell type. Split-open tubules were loaded in the experimental chamber with BCECF (5 μM) from both the basolateral and luminal sides at room temperature for 15 min, after which they were superfused with sodium Ringer solution (solution 1) at 37°C for at least 15 min prior to the start of the experiment. Calibration of the emitted signal from each cell studied was performed at the end of each experiment. Extracellular pH was varied from 6.8 to 7.8 (solutions a and b) in the presence of the K/H exchanger nigericin (10 μM) in 145 mM K according to the method of Thomas et al. (29). This pH range is in the linear portion of the calibration curve for this dye, as we and others have previously shown (4, 25). The experimentally determined fluorescence ratios were then transformed to pH by the imaging software (Metafluor) on a pixel-by-pixel basis. Each tubule was calibrated in this manner.
Under epifluorescence, ICs were visually identified and distinguished from the neighboring principal cells based on differential loading of BCECF (25, 37). Figure 1 is a typical example of the loading pattern of BCECF after being exposed to 5 μM of the dye for 15 min. Figure 1 A is a transmitted light image of a split-open tubule, and Fig.1 B is the fluorescence image of the same tubule. In the transmitted light image, the dark, cusp-shaped cells are visually identified as ICs, and the cells interspersed amongst them with the pronounced nuclei are the principal cells. The corresponding fluorescence image demonstrates that the cells with the brightest fluorescence correspond to those cells visually identified as the ICs. This difference in fluorescence presumably reflects a far greater accumulation of the cleaved form of BCECF-AM, due to higher intrinsic esterase activity in the ICs compared with the principal cells.
Bloods and Urines
Blood samples were obtained from some animals either by cardiac puncture or ocular bleeding and analyzed for pH and , Na+, K+, and Cl−. Blood samples were analyzed on two different blood microsystems, which may account for the variability observed. Urine was collected over a 24-h period under mineral oil using specially designed metabolic cages and analyzed for pH, Na+, K+, and Cl−.
Results are expressed as means ± SE, wheren refers to the number of cells individually analyzed. The mean responses of all of the ICs (±SE) studied in each experimental group are presented in Figs. 4, 7, and 8and summarized in Table 3. Representative experimental traces from individual tubules are shown in Figs. 2, 3, 5, and 6. These traces depict the mean intracellular pH (pHi) response (±SE) of the ICs studied for that individual tubule. Anywhere from two to five ICs were studied in any given tubule preparation, depending on the width of the tubule. The NH4Cl pulse protocol was performed only once on each tubule followed by the in situ nigericin calibration. Significant differences were determined by ANOVA. Significance was asserted if P< 0.05.
Effects of a Low-NaCl Diet on Acid-Base and Electrolyte Balance
Table 2 lists the results of blood and urine analysis from control and NaCl-deficient rats. The blood values for the control and NaCl-deficient rats were quite similar except for a slight decrease in plasma Na+levels. Blood pH and plasma K+values were not different between the two groups. Urine pH was the same for both groups. As expected, urine Na+ and Cl− concentrations were much lower in the experimental group compared with controls.
H-K-ATPase function was assayed as the rate of K-dependent intracellular alkalinization in response to an acute acid load of 10 mM . As part of this protocol and in all of the experiments presented in this study, bafilomycin A1 (100 nM), a vacuolar H-ATPase inhibitor (2), was added to all solutions from the NH4Cl acid pulse solution (solution 2) through the end of the experimental protocol at a concentration used by others to inhibit H-ATPase activity in rat late distal tubule (17, 18, 34).
K-Dependent Intracellular Alkalinization in Control ICs
An example of the response to the acid pulse protocol by three ICs from a single tubule from a control rat is shown in Fig.2. Each representative point is the mean ± SE of the three ICs simultaneously studied from this tubule. The pHi value shown on the ordinate was determined from the 490/440 fluorescence ratio and the in situ calibration performed on this tubule. The solution changes are depicted at the top of Fig. 2. As shown, upon removal of , Na+, and K+ (solution 3), the pHi fell more than 1 pH unit from the initial pHi in all of the cells monitored. As expected, no pHi recovery was observed, because of the continued presence of bafilomycin A1 in the Na- and K-free superfusing solution. Readdition of 5 mM K to the superfusate resulted in a small and partial K-dependent pHi recovery at a mean rate of 0.045 ± 0.01 pH units/min. The K-dependent pHi recovery rate from all of the ICs studied in control tubules averaged 0.07 ± 0.01 pH units/min, which is similar to rates reported in earlier and similar studies on control rabbit ICs (5, 26, 27). Slopes of the K-dependent (and Na-dependent) intracellular alkalinization rates were calculated from the point where recovery had actually started to the leveling off point, as illustrated in Fig. 2. The time lag sometimes observed between solution changes and the cellular response reflects the time it may take for the solution to exit the Hamilton valve, become warm, and then replace the preexisting solution in the chamber. The solution marker on the trace reflects the moment right after the valve is manually turned to the appropriate solution. A summary of the responses of all of the ICs from control tubules is shown in Table 3. Under control conditions, H/K exchange raised pHi by ∼0.2 pH units to a mean pHi of 6.72 ± 0.04.
Introducing Na back to the superfusate resulted in additional recovery at a rate of 0.54 ± 0.12 pH units/min. This Na-dependent recovery reflects basolateral Na/H exchange similar to that described in rabbit CCT (27, 36).
Response to Sch-28080
To test whether the K-dependent intracellular alkalinization observed in control tubules was due to an H-K-ATPase, the same protocol was carried out in the presence of the gastric H-K-ATPase inhibitor Sch-28080 at a concentration known to completely inhibit the gastric isoform (10 μM) (33). Figure 3 is an experimental trace of the means ± SE from three ICs in the same control tubule exposed to the NH4Cl protocol. As shown in Fig.3, there was virtually no K-dependent pHi recovery in the presence of the imidazopyridine inhibitor, as indicated by the low rate of pHi recovery 0.01 ± 0.01 pH units/min. As shown in Table 3, the mean overall rate of the Sch-28080-insensitive K-dependent pHi recovery in all of the ICs from control tubules was 0.03 ± 0.01 pH units/min, with the Sch-28080-sensitive component being ∼0.04 pH units/min.
Sch-28080 did not appear to affect the rate of Na/H exchange. As shown in Fig. 3, readdition of Na to the superfusate resulted in a rapid intracellular alkalinization response back to the initial pHi at a rate of 0.85 ± 0.10 pH units/min.
To test whether the Sch-28080-insensitive component of the K-dependent pHi recovery rate could be due to a ouabain-sensitive isoform of H-K-ATPase, the acid pulse protocol described above was performed in the presence of 1 mM ouabain. TheK i for ouabain inhibition of the α-subunit of the putative colonic H-K-ATPase expressed in oocytes in the presence of 5 mM extracellular K was close to 1 mM (6).
Figure 4 compares the K-dependent pHi recovery rates in the absence and presence of ouabain and/or Sch-28080. Only the addition of Sch-28080 alone or with ouabain significantly inhibited the K-dependent pHi recovery rate (P < 0.01). Ouabain alone did not have any affect on the K-dependent pHi recovery rate. Also, the rate of K-dependent pHi recovery was the same in the presence of Sch-28080 and Sch-28080 and ouabain combined (see Table 3). Taken together, these data demonstrate that a Sch-28080-sensitive H-K-ATPase predominates as the primary H/K exchanger in ICs under control conditions. A summary of all of these results is shown in Table 3.
Effect of Low-NaCl Diet
Having demonstrated H/K exchange in individual ICs from control tubules, we next tested the hypothesis that NaCl restriction induces H/K exchange in ICs of the CCT. Experiments similar to those described above were performed on tubules dissected from rats maintained on a NaCl-restricted diet.
Figure 5 is representative record from one experiment of four ICs identified in a tubule from a rat maintained on a low-NaCl diet. Bafilomycin A1(100 nM) was present from the acid pulse through the end of the protocol. As shown in Fig. 5, upon removal of , Na+, and K+, the pHi fell from an initial mean pHi of ∼7.5 to 6.7. No pHi recovery was observed in the absence of extracellular Na and K. With readdition of K to the superfusate, partial pHi recovery occurred, returning the pHi to a mean of 6.85. For the four ICs in this tubule, the K-dependent rate of alkalinization averaged 0.11 ± 0.04 pH units/min. As shown in Table3, for all of the ICs studied under low-NaCl conditions, the mean K-dependent pHi recovery rate was roughly double that observed under control conditions [0.13 ± 0.02 pH units/min for low NaCl (n = 35 cells) vs. 0.07 ± 0.01 pH units/min for control (n = 38 cells)].
As shown in Fig. 5, introducing Na back to the superfusate resulted in additional recovery for these four ICs at a mean rate of 0.15 ± 0.03 pH units/min back to the initial pHi. The mean Na-dependent pHi recovery rate for 35 ICs studied with low NaCl was 0.28 ± 0.04 pH units/min.
Response to Sch-28080 and Ouabain in NaCl-Deficient Rats
To test whether the increased rate of K-dependent pHi recovery observed in the low-NaCl ICs was due to stimulation of the Sch-28080-sensitive H/K exchanger, pHi recovery was monitored in the presence of 10 μM Sch-28080. A representative result from one experiment on a tubule is shown in Fig.6. As shown, readdition of 5 mM K in the presence of Sch-28080 partially inhibited K-dependent pHi recovery with pHi going from 6.6 to 6.75. The rate of K-dependent intracellular alkalinization in the presence of Sch-28080 for the two ICs in this tubule was 0.07 pH units/min. In all of the ICs studied in the presence of Sch-28080 from low-NaCl rats, the average rate of K-dependent pHirecovery was 0.05 ± 0.02 pH units/min. The rate of this recovery is less than one-half of that observed in ICs from NaCl-deficient rats without inhibitor (0.13 ± 0.02 pH units/min) (Table 3). These data suggested that an additional Sch-28080-insensitive component was contributing to the rate of K-dependent intracellular alkalinization in the ICs from low-NaCl rats.
To examine the possibility that NaCl restriction induces a ouabain-sensitive isoform of H-K-ATPase in CCT, we compared rates of K-dependent intracellular alkalinization in the absence and presence of ouabain and Sch-28080.
Figure 7 compares the mean K-dependent pHi recovery rates for all of the low-NaCl ICs studied in the absence and presence of Sch-28080 and/or ouabain. The K-dependent rate of intracellular alkalinization is greatest in the low-NaCl ICs (roughly double that measured in control ICs; Fig. 4 and Table 3). Addition of 1 mM ouabain to the superfusate, a concentration that did not have any effect on the K-dependent pHi recovery rate in control ICs (Fig. 4), significantly inhibited the K-dependent pHi recovery rate from 0.13 ± 0.02 to 0.06 ± 0.01 pH units/min (P < 0.01). This inhibition of the K-dependent pHi recovery rate was similar to that observed in the presence of Sch-28080 (0.05 ± 0.02 pH units/min), which was also significantly lower than control rates (P < 0.001). Adding both Sch-28080 and ouabain together to the superfusate totally inhibited the rate of K-dependent pHi recovery (0.01 ± 0.01 pH units/min), suggesting that both isoforms of H-K-ATPase contribute to K-dependent pHirecovery in ICs from NaCl-restricted rats.
Effect of Aldosterone
Because chronically maintaining rats on a NaCl-deficient diet results in elevated endogenous plasma aldosterone levels, we speculated that aldosterone may be the mediator responsible for inducing the H-K-ATPase exchange rate in ICs from low-NaCl rats. To test this idea, a group of rats was implanted with osmotic minipumps, which continuously administered aldosterone for 12–14 days at levels comparable to that measured in our previous study on low-NaCl rats (21). In our present study, the amount of aldosterone delivered by the pumps was 250 μg ⋅ kg body wt−1 ⋅ day−1for 12–14 days. This was the same amount of aldosterone put in the pumps used in our previous study (21), which produced plasma aldosterone levels of 750 ng/dl in identical rats. Figure8 compares the K-dependent pHi recovery rates in the ICs from the minipump rats with control ICs and ICs from rats maintained on the low-NaCl diet. The rate of K-dependent pHi recovery measured in ICs from the rats infused with aldosterone was comparable to that measured in control rats (0.06 ± 0.01 vs. 0.07 ± 0.01 pH units/min). These results demonstrate that aldosterone alone is not responsible for the enhanced rate of K-dependent intracellular alkalinization observed in ICs from low-NaCl animals. Some other as yet unidentified mechanism is responsible for this induced change.
The BCECF-loaded, split tubule preparation was used in combination with dual-excitation digital imaging to study H/K exchange in ICs in response to an imposed acidosis. ICs were visually identified under epi-illumination by their much brighter appearance compared with the neighboring principal cells. Our results indicate that maintaining rats on a NaCl-restricted diet induces a ouabain-sensitive H/K exchanger in the ICs that is not observed under control conditions. With chronic NaCl deficiency, both the Sch-28080- and the ouabain-sensitive H/K exchangers contribute equally to the K-dependent pHi recovery rate after an acute acid pulse. The net effect is an increased rate of K-dependent pHi recovery from an acid load in the low-NaCl ICs compared with control ICs.
In both control and low-NaCl ICs, the K-dependent intracellular alkalinization in response to the acid load resulted in only a partial recovery of the pHi (Figs. 2 and5). The magnitude of this pHirecovery was not different in the low-NaCl ICs compared with controls (0.22 ± 0.02 for low-NaCl ICs vs. 0.19 ± 0.03 pH units for control ICs). This type of response has also been observed in ICs from rabbits under control and chronically acidotic conditions (26, 27). In ICs from newborn rabbits, however, the K-dependent pHi recovery was more complete, returning the pHi back to the basal value (5). This difference in response between the ICs from newborns and adults indicates that the H-K-ATPase is capable of functioning at pHi > 7.0 but there is some underlying factor in mature animals under the conditions of these studies that limits the function of the H-K-ATPase.
In all of the experiments, readdition of Na to the superfusate results in additional pHi recovery; however, the rates are different between the controls and low-NaCl ICs. The ICs from control rats had much higher rates than ICs from low-NaCl rats and aldosterone minipump rats [0.44 ± 0.04 (n = 38 ICs) for control, vs. 0.28 ± 0.04 pH units/min (n = 35 ICs) for low NaCl, vs. 0.31 ± 0.04 pH units/min (n = 32 ICs) for aldosterone minipump]. This difference is also evident by comparing the Na-dependent slope in a control tubule (Fig. 2) with that measured in ICs from a low-NaCl rat tubule (Fig. 5). We did not explore the reasons for these differences. They may be related to the higher plasma aldosterone levels in the low-NaCl and aldosterone minipump rats compared with controls.
The isoform or variant of the ouabain-sensitive H-K-ATPase, induced in the ICs with NaCl restriction, remains to be determined. Although the pharmacological profile of this exchanger has not yet been fully characterized, 1 mM ouabain appeared to inhibit H/K exchange. We do not know whether this dose of ouabain was maximal. It also remains to be determined whether this H-K-ATPase is inhibitable by a higher concentration of Sch-28080 than that used in this study (10 μM). Based on enzymatic assays, it was reported recently that another H-K-ATPase isoform may be expressed in rat CCT with K depletion (3). The enzymatic activity for this H-K-ATPase was inhibited with 1 mM ouabain and 100 μM Sch-28080.
Our finding suggesting that two isoforms of H-K-ATPase function in ICs of NaCl-deprived rats is consistent with mRNA expression studies. The mRNA for the α-subunit of gastric H-K-ATPase has been detected in CCT from normal rats (1). Similarly, mRNA for the α-subunit of the putative distal colonic H-K-ATPase has also been detected in this nephron segment from control rats (31). By means of a modified immunoprecipitation technique, protein levels associated with the α-subunit of colonic H-K-ATPase and gastric H-K-ATPase have been measured in kidneys from control rats (15). Furthermore, this study reported the finding that a K-deficient diet increased the abundance of protein for the colonic isoform of H-K-ATPase, whereas that of the gastric isoform remained at control levels (15). Differential expression of H-K-ATPase isoforms was also observed at the mRNA level in kidney cortex from NaCl deprived rats (35). With Northern hybridization techniques, a fourfold increase in the mRNA level for the α-subunit of colonic H-K-ATPase was found with NaCl restriction, but the level of message for the α-subunit of the gastric isoform remained at control levels (35). This is similar to the results of Sangan et al. (23), who also saw an increase in the mRNA level for colonic H-K-ATPase in renal cortex with NaCl depletion.
The experiments presented in this study also demonstrate that aldosterone is not responsible for the enhanced H-K-ATPase activity in ICs from NaCl-depleted rats (Fig. 8). This finding is in agreement with those of Eiam-Ong (9), who demonstrated that Sch-28080-sensitive H-K-ATPase activity in the CCT was not altered with increased plasma aldosterone. More recently, it was demonstrated that the level of mRNA expression of the putative colonic K-ATPase was not altered in rat kidney with elevated plasma aldosterone (14). These results in kidney are different from the findings in distal colon, where it has been shown that aldosterone administration via osmotic minipump produced enhanced electroneutral K+reabsorption (32) believed to be mediated via a K-ATPase (14).
We are left to consider the physiological significance of increased H-K-ATPase function under conditions of NaCl restriction. Under normal conditions in control rats, the reabsorption of Na+ is minimal in the collecting duct as a result of a dearth of Na channels in the principal cells; however, under conditions of NaCl restriction with the resultant elevated plasma aldosterone levels, there is an abundance of apical Na channels in the principal cells (12, 22). The higher density of Na channels as well as increased channel activity under these conditions leads to increased reabsorption of Na+ from the tubular fluid as demonstrated in isolated microperfused rabbit and rat CCTs (24, 30). It is believed that the increase in the principal cell apical membrane Na+ permeability results in a more favorable electrical driving force for secretion of K+ and leads to a coupling between Na+ reabsorption and K+ secretion (22, 28). NaCl restriction is also known to increase Cl− reabsorption in the collecting duct independent of plasma Cl− concentration, filtered Cl− load, or Cl− delivery to the collecting duct. (10).
We speculate that the enhanced H/K exchange observed with NaCl restriction may serve to reabsorb the K+ secreted in the neighboring principal cell. This would necessitate that H-K-ATPase be located in apical membrane of the ICs. We and others have previously demonstrated apical localization of H-K-ATPase in β-type ICs of rabbit CCT (26, 38). Although we were not able to differentiate between the IC subtypes in rat CCT, ultrastructural analysis of kidney from control rat has shown a predominance of β-type ICs in the CCT (19). In our model of the β-type IC, the apical membrane contains H-K-ATPase and the Cl−/ exchanger. Under conditions of NaCl restriction, enhanced reabsorption of Na+ via the principal cells leads to increased excretion of K+into the tubular lumen. This secreted K+ could then be reabsorbed via the H-K-ATPase in the neighboring ICs, concomitant with exchange of luminal Cl− for . The net effect of these coordinated processes is reabsorption of Na+, K+ and Cl− and the secretion of H+ and .
In conclusion, we have demonstrated increased H-K-ATPase function in ICs of rat CCT with chronic NaCl deficiency. Based on inhibition by Sch-28080 and ouabain, we conclude that this functional H/K exchange maybe due to two distinct isoforms. The signal responsible for the increased H/K exchange appears not to be related to acid-base state, plasma K+ balance, or plasma aldosterone levels. In NaCl depletion, the increased H/K exchange in addition to increased Na+reabsorption and K+ secretion yields net Na+ for H+ exchange. This in parallel with apical Cl−/ exchange in the β-type ICs, may serve to promote net NaCl reabsorption by the CCT rather than H+ for K+ exchange.
We thank Drs. E. E. Windhager and A. Weinstein for critical readings of this manuscript.
Address for reprint requests: R. B. Silver, Dept. of Physiology, Cornell Univ. Medical College, 1300 York Ave., New York, New York 10021.
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-45828 and DK-11489 and by the Underhill and Wild Wings Foundations (to R. B. Silver).
- Copyright © 1998 the American Physiological Society