INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

PREVIOUS STUDIES from our laboratory have identified two H⁺-K⁺-ATPase -isoforms in rat whole kidney by Northern analysis: the gastric or the HK₁ isoform, and the colonic or the HK₂ isoform (13, 14). Importantly, however, in response to chronic hypokalemia, there was a significant and selective increase in HK₂ mRNA abundance; HK₁, in contrast, did not respond to hypokalemia. A similar increase in HK₂ mRNA abundance in response to chronic dietary potassium deprivation was observed in separate studies by our laboratory in the distal colon (9).

HK₁ mRNA has been localized to the cortical collecting duct by in situ hybridization techniques (1, 2, 34). Furthermore, using semiquantitative techniques, the abundance of HK₁ mRNA has been observed to increase selectively with chronic dietary potassium depletion in this same nephron segment (2). Conversely, in response to chronic hypokalemia, the abundance of HK₂ mRNA increased selectively in the medullary collecting duct (MCD) but not in the cortical collecting duct (3). Recently, Kraut and associates (21) used immunoprecipitation techniques to examine the abundance of HK₁ and HK₂ protein in a rat kidney microsomal fraction. This study reported that both proteins are expressed in the kidney of control rats. Moreover, in this same study, a low-potassium diet induced a marked increase in HK₂ protein abundance but had no impact on the abundance of HK₁.

In addition, Jaisser and associates (18) have reported, using the RNase protection assay, that HK₂ mRNA was expressed in both kidney and distal colon. In contrast to our findings, they have reported that the renal HK₂ and colonic HK₂ were differentially regulated; i.e., chronic hypokalemia augmented the abundance of HK₂ mRNA in kidney but not in distal colon. Other investigators have also observed that chronic hypokalemia increased the abundance of mRNA in the kidney but not in distal colon (29). However, although hypokalemia had no effect on HK₂ expression in distal colon, dexamethasone or aldosterone administration to adrenalectomized rats increased the abundance of HK₂ mRNA in both organs (18). In the study by Binder and colleagues (29), it was reported that the abundance of HK₂ protein in rat kidney, as recognized by a polyclonal antibody against the amino terminal of HK₂, increased after chronic K⁺ depletion. Nevertheless, a similar regulatory response was not observed in the distal colon, where the levels of mRNA and protein were upregulated only in response to dietary sodium depletion (which was assumed to be synonymous with an increase in aldosterone elaboration) (29).

Based on the existing uncertainties in this area, we raised antibodies that were highly specific for HK₁ and HK₂, to elucidate more clearly the distribution and regulatory response of these two proteins in kidney and distal colon. The results demonstrate that HK₂ is upregulated selectively in the renal medulla, but not in renal cortex, in response to chronic dietary K⁺ depletion. There was no response to chronic metabolic acidosis (CMA). Moreover, HK₁ protein was not detected in either region of the kidney. Furthermore, while HK₂ was readily detectable in the distal colon of control rats, there was no variation in protein abundance in response to either chronic hypokalemia or CMA.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Experimental animals. Chronic hypokalemia (LK) was generated in male Sprague-Dawley rats (135-175 g) by dietary K⁺ restriction. Rats in the LK group received tap water to drink and were maintained for 14 days on a customized vitamin-fortified nominally K⁺-free diet that contained 0.001 nmol/g of potassium (lot 960189; ICN Biochemicals, Cleveland, OH). Control rats were pair fed each day the same diet, to which was added KCl 0.03 mmol/g (no. P5411; ICN Biochemicals), in amounts identical to that consumed by the LK rats, and tap water was consumed ad libitum. CMA was generated and maintained as established by our laboratory previously (14). Animals in this group were fed the same diet as controls, but instead of tap water, these rats drank 0.28 M NH₄Cl for 4 days. All dietary modifications were well tolerated, and produced stable and reproducible models of either dietary K⁺ depletion or CMA. The rats were killed in sets of three (one for each experimental condition), and the proteins were prepared in corresponding sets to minimize bias during the preparation.

Preparation of plasma membranes. To prepare plasma membranes (26), either specific regions of the kidney or the distal colon (as appropriate) were homogenized (1.0 g of tissue; using a Brinkmann Polytron, model PT 10/35) in the presence of 10 mM Tris · HCl, pH 8.0, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride (PMSF), 3 mM benzamidine, and 1 µg/ml soybean trypsin inhibitor (buffer A) containing 27% sucrose (wt/vol). Nuclei were removed by centrifugation at 2,000 g for 4 min at 4°C, and the supernatant was applied to the top of 45% (wt/wt) sucrose in buffer A and was centrifuged at 200,000 g for 45 min at 4°C. The membranes in the interphase 27/45% sucrose were diluted in buffer A and collected by centrifugation at 25,000 g. The final protein concentration was measured using the Lowry method (24).

Preparation of apical membranes. Apical membranes from distal colon were prepared as described by Aronson (4). Distal colon (1 g in 10 ml buffer B: 300 mM mannitol, 1 mM Tris-HEPES, pH 7.5, 1 mM PMSF, 3 mM benzamidine, 1 µg/ml soybean trypsin inhibitor) was homogenized with a Polytron as described above. The homogenization was performed at 4°C. The particulate material was removed by centrifugation for 2 min at 200 g. To the supernatant, 1 M MgSO₄ was added to a final concentration of 10 mM MgSO₄. The sample was placed on ice for 15 min and was shaken intermittently. The aggregated material was removed by centrifugation for 12 min at 2,500 g at 4°C, and the apical membranes from the supernatant were collected by centrifugation at 27,000 g for 20 min at 4°C. The top of the pellet was resuspended in 5 ml buffer B containing 10 mM MgSO₄. After removal of the aggregated material at 3,100 g for 12 min at 4°C, the membranes were collected by centrifugation at 27,000 g for 20 min at 4°C. The final membranes were resuspended in 10 mM mannitol and 2 mM Tris · HCl, pH 7.5, and the protein concentration was measured using the method of Lowry et al. (24).

Immunoblots. Plasma or apical membranes, prepared as described above, were separated on 10% SDS-PAGE (except when indicated) as described previously by our laboratory (17). The proteins were transferred overnight at 30 V to a nitrocellulose membrane (no. BA85; Schleicher and Schuell) in the presence of Tris-glycine buffer [25 mM Trizma base, 250 mM glycine, and 20% (vol/vol) methanol]. The nonspecific binding sites of the nitrocellulose membrane were blocked in the presence of 5% nonfat dry milk in PBS-Tween (10 mM sodium phosphate, pH 7.5, 150 mM NaCl, and 0.05% Tween-20). This was followed by 2 h incubation with the primary antibody diluted 1:1,000-10,000 in PBS-Tween. After extensive washing, the membranes were incubated with a peroxidase-bound donkey-anti-rabbit IgG, and the bands of interest were detected using an enhanced chemiluminescence system (ECL, Amersham no. RPN2108) following the manufacturer's instructions.

Anti-HK₁ antibody. A synthetic peptide (TLEDPRDPRHL) (30) that extends from amino acid 503 to 513 of the published rat HK₁ sequence was synthesized (Bethyl Laboratories, Montgomery, TX). Two rabbits were injected and bled according to standard protocols. This region was chosen because this HK₁ peptide is distinct from other X⁺-K⁺-ATPases. The maximum identity was with the ₃-Na⁺-K⁺-ATPase (54% identity), whereas HK₂ and ₁- and ₂-Na⁺-K⁺-ATPases were significantly more divergent. Moreover, the selected peptide does not contain a consensus sequence for secondary modifications as assessed by the Motifs program of the GCG package (Wisconsin Sequence Analysis Package).

Anti-HK₂ antibody. A synthetic peptide (TPEQLDELLTNYQ) (22) that extends from amino acid 686 to 698 of the published HK₂ sequence (12) was synthesized (Genosys Biotechnology, The Woodlands, TX). Two rabbits were injected and bled according to standard protocols. Maximum identity was with the human-ATP1AL1 (69% identity), which is known to be expressed in brain, skin, and kidney (19). Maximum divergence was with rat HK₁ (23% identity). This peptide does not contain a consensus sequence for secondary modification as assessed by the Motifs program of the GCG package (Wisconsin Sequence Analysis Package).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Specificity of anti-HK₂. Two different assays were performed to determine the specificity and sensitivity of the anti-HK₂ antibody. Specificity was established using membranes prepared from different tissues, as follows: the stomach, which is enriched in HK₁; the kidney, which is enriched in ₁-Na⁺-K⁺-ATPase; and the distal colon, which is enriched in HK₂. The results of one of these experiments are displayed in Fig. 1. The anti-HK₂ antibody hybridized at a mobility near 100 kDa (the expected mobility for HK₂) in membranes prepared from distal colon (Fig. 1, left). However, the antibody did not hybridize to plasma membranes prepared from control rat whole kidney, renal cortex, renal outer medulla, renal inner medulla, or stomach, even when as much as 100 µg protein was applied to the gel. A band was also detected in rat brain plasma membranes in the region of 100 kDa. The 100-kDa band in distal colon and brain plasma membranes was completely blocked by preincubation of the immune serum with the immunizing peptide (200 µM for 1 h at 4°C) (Fig. 1, right).

View larger version (36K): [in this window] [in a new window]   Fig. 1.   Specificity of anti-HK2 antibody. Left: plasma membranes prepared from different tissues from control rats were resolved on SDS-PAGE, transferred to a nitrocellulose membrane, and hybridized against the anti-HK2 antibody (dilution 1:1,000). The anti-HK2 antibody hybridized to distal colon and brain plasma membranes only. Right: when immune serum was preincubated for 1 h at 4°C with the immunizing peptide (200 µM), the signal was completely blocked in both membranes.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   HK2 localizes to the apical membrane of the distal colon. Distal colon plasma (lane 1) or apical (lane 2) membranes (5 µg each) were separated on 7% SDS-PAGE, transferred to a nitrocellulose membrane, and probed with the anti-HK2 antibody (dilution 1:1,000). Lane 3: distal colon plasma membranes (10 µg) were probed with a mixture of anti-HK2 antibody (1:1,000) and anti-1-Na+-K+-ATPase antibody (1:1,000). Apical membranes are enriched in HK2 compared with the total plasma membranes. (Note the single sharp band; see text.)

HK₂ protein is upregulated in the rat medulla by chronic dietary K⁺ depletion. As described above, the levels of HK₂ were undetectable in membranes prepared from renal medulla prepared from control rats (Fig. 1 and Fig. 3, middle). This observation suggests that HK₂ was expressed at very low levels.

View larger version (38K): [in this window] [in a new window]   Fig. 3.   Selective response of HK2 to chronic hypokalemia (LK) in the renal medulla. Left: recognition of renal medullary membranes (10 µg) by anti-1-Na+-K+-ATPase antibody (LEAVE) (dilution 1:1,000). Middle: 100 µg of the same membranes as left hybridized against the anti-HK2 antibody (dilution 1:1,000). Right: 100 µg of the same membranes hybridized against the anti-HK1 antibody (dilution 1:1,000). Only the region of 100 kDa is shown. Densitometric analysis of 1-Na+-K+-ATPase was as follows (in arbitrary units): control = 2.4 ± 0.3; LK = 3.2 ± 0.2; chronic metabolic acidosis (acidosis) = 3.7 ± 0.3. Densitometry analysis of HK2 was as follows: control = not detectable; LK = 0.5 ± 0.1; acidosis = not detectable. None of the differences were statistically significant. Each group contained 4-6 rats.

View larger version (42K): [in this window] [in a new window]   Fig. 4.   HK2 is not upregulated by chronic hypokalemia or metabolic acidosis in the renal cortex. Immunoblots were performed as described in Fig. 3. Densitometric analysis of 1-Na+-K+-ATPase was as follows (in arbitrary units): control = 1.4 ± 0.2; LK = 2.0 ± 0.5; acidosis = 1.7 ± 0. These differences were not significant. Each group contained 4-6 rats.

View larger version (39K): [in this window] [in a new window]   Fig. 5.   HK2 is not upregulated by chronic hypokalemia or metabolic acidosis in the distal colon. Immunoblots were performed as described in Fig. 3, except that 10 µg of membranes were used in the immunoblots with the anti-HK2 antibody (middle). Short exposure of the film is shown to maximize the linearity of the assay. Densitometric analysis of 1-Na+-K+-ATPase was as follows (in arbitrary units): control = 2.5 ± 0.6; LK = 2.1 ± 0.3; acidosis = 2.1 ± 0.1. Densitometric analysis of HK2: control = 1.1 ± 0.2; LK = 1.3 ± 0.2; acidosis = 0.7 ± 0.2. None of these differences were significant. Each group consisted of 4-6 rats.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The distal colon and the cortical and medullary collecting tubules contribute significantly to systemic potassium homeostasis (3, 9, 21). Functional studies have indicated that an ATP-dependent H⁺/K⁺ exchanger resides in mammalian collecting duct segments (15, 19, 33, 34) where it catalyzes H⁺ secretion and K⁺ absorption. Studies using hybridization techniques have demonstrated that mRNA for both HK₁ and HK₂ are expressed in whole kidney and in cortical collecting ducts, outer MCD (OMCD) and inner MCD (IMCD) derived from chronically hypokalemic rats or rabbits (1, 2, 14, 18, 35). Therefore, these two H⁺-K⁺-ATPase isoforms have been proposed as critical to the regulation of K⁺ reabsorption by the kidney (12, 19). It has been reported that HK₁ mRNA abundance was increased in the rat cortical collecting duct after chronic dietary K⁺ depletion (1). Studies in whole kidney from our laboratory, however, revealed that chronic hypokalemia was associated with a selective increase in HK₂, not HK₁ mRNA abundance (14). If a selective increase in HK₁ mRNA was confined to the cortex, then it seems reasonable to expect that, since the cortical fraction would account for the majority of whole kidney total RNA, a similar response should be anticipated in whole kidney as well. We proposed, therefore, that the apparent regulatory response to chronic hypokalemia by the H⁺-K⁺-ATPase in kidney was specific for HK₂.

However, until very recently (21), there were no studies available that investigated either HK₁ or HK₂ protein in kidney during chronic dietary K⁺ depletion. Kraut and associates (21) have observed that although both HK₁ and HK₂ protein were detectable at very low levels in microsomes from rat whole kidney (using a modified immunoprecipitation technique), only HK₂ protein increased in response to chronic dietary K⁺ depletion. Thus the findings by Kraut and associates (21) for HK₂ protein are compatible with our previous report for HK₂ mRNA. Moreover, the 6- to 10-fold increase in HK₂ mRNA abundance in the rat whole kidney following chronic dietary K⁺ depletion, reported previously (14), is compatible with the magnitude of the increase in HK₂ protein observed in the present study.

A unique finding of the present study is that for the first time, the increased HK₂ protein abundance can be localized specifically to the renal medulla, not the renal cortex per se (Fig. 3 vs. Fig. 4). Upregulation of HK₂ mRNA following chronic dietary K⁺ depletion was reported by Ahn et al. (3) using in situ hybridization techniques in the MCD, but not in the cortical collecting duct (3). When these findings are considered together, it seems reasonable to conclude that the parallel increase in HK₂ mRNA and protein with chronic hypokalemia is specifically localized to the MCD.

Kone and Higham (20) recently reported an increase in HK_2b mRNA in renal medulla in response to chronic hypokalemia. However, our anti-HK₂ antibody, which should detect both HK₂ (migrating at 115 kDa) and the splice variant, HK_2b (migrating at 102.5 kDa), detected only HK₂. Thus HK₂, not HK_2b, appears to be the major H⁺-K⁺-ATPase transporter that participates in the regulation of K⁺ absorption in the renal medulla, in response to chronic hypokalemia. Moreover, these findings are in agreement with recent results from Kraut and associates (21). In that study, where larger quantities of renal microsomes were probed with a similar antibody, a protein with a mobility approximating that of the newly reported splice variant (HK_2b) was not detected in either control or hypokalemic rats. In addition, the antibody employed by Binder (29) against the amino terminal of HK₂, which should not detect HK_2b, also revealed an increase in HK₂ protein abundance during chronic hypokalemia. When taken together these studies do not predict a major role for HK_2b protein in the regulatory response to chronic hypokalemia by the MCD. Furthermore, this same logic would apply to the distal colon, where HK_2b mRNA has been reported to be expressed abundantly (20). That we were unable to detect HK_2b protein in distal colon plasma membranes suggests that the protein is not expressed in this location as well.

When HK₂ has been expressed in a heterologous system, in studies by our laboratory, it has been demonstrated that a -subunit was required for activity (to support both Rb⁺/K⁺ uptake and H⁺ extrusion) (8, 10). Moreover, using this functional assay in oocytes, HK₂ was completely insensitive to Sch-28080 but relatively sensitive to ouabain (IC₅₀ = 400-500 µM). Nevertheless, recent transport studies from our laboratory (33) have demonstrated an increase in H⁺-K⁺-ATPase-mediated bicarbonate absorption in the IMCD following short-term dietary K⁺ depletion. In response to chronic hypokalemia, there was an increase in both total and Sch-28080-sensitive total bicarbonate flux (J_tCO2) in the terminal IMCD (tIMCD) perfused in vitro. Moreover, a high concentration of ouabain in the perfusion solution (5 mM) decreased J_tCO2 both independently and additively to the inhibition by Sch-28080 (33). Thus it appeared that in the rat tIMCD, the increase in bicarbonate absorption in response to chronic hypokalemia could be attributed to both HK₁ and HK₂. Moreover, intracellular pH (pH_i) recovery studies in mouse OMCD inner stripe cells (OMCD-1 cells) have demonstrated a similar adaptive response by "the" H⁺-K⁺-ATPase to a chronic reduction in K⁺ concentration in cell culture media (15). In OMCD the increase in dpH/dt in response to simulated hypokalemia (15) was fully inhibited by low concentrations (10 µM) of Sch-28080, suggesting upregulation of H⁺ secretion by the HK₁, not HK₂.

The reason for the apparent discrepancy between functional data in isolated perfused tubules and OMCD cells in culture compared with functional data obtained from heterologous expression systems is not immediately apparent. Several explanations for these differences should be considered if the findings using these techniques are to be reconciled. These possibilities include but are not limited to 1) adaptation to chronic hypokalemia by an H⁺-K⁺-ATPase (or splice variant) in kidney which has not yet been identified; 2) secondary modifications in HK₂ in the environment of the epithelial cell that alter its affinity for Sch-28080; or 3) posttranscriptional or posttranslational modification, which could account for such regulatory diversity. 4) In addition, although not evident in heterologous systems where expression of HK₂ with known -subunits (HK_G, ₁) results in identical function, the ₁-Na⁺-K⁺-ATPase in vivo could alter the pharmacological sensitivities from that observed in X. laevis oocytes. 5) Another possible explanation is that the increase in Sch-28080-sensitive acidification in the tIMCD occurs through increased trafficking of HK₁ between the plasma membrane and the cytosol. In the stomach, regulation of HK₁-mediated H⁺ secretion occurs through trafficking of HK₁ protein between the plasma membrane and intracellular vesicles (11). In the kidney, it is possible that the increase in H⁺-K⁺-ATPase activity observed following dietary K⁺ restriction occurs through increased insertion of HK₁ into the apical membrane to a level not detectable by the immunoblot used in the present study. Because a single explanation has not been identified, in view of these uncertainties, until additional studies can be performed, the pharmacological profiles ascribed to the HK₂ in transport studies must be viewed with caution.

Western blot analysis of kidney samples under control conditions has revealed a protein of similar mobility to the gastric H⁺-K⁺-ATPase -subunit in some studies (7) but not in others (6). The recent report of Kraut and associates (21), mentioned above, also detected HK₁ protein in enriched microsomes harvested from the whole kidney of rats under conditions of normal acid-base and potassium balance. Our antibody did not recognize HK₁ protein in either cortex or medulla of the kidney from control rats. We attribute this difference in results to variation in the methodology employed in the two studies. Based on the description of the technique employed by Kraut et al. (21), there appears to be no limit in the quantity of microsomes that could be used in the immunoprecipitation experiments. In the immunoblots used in our experiments, resolution of the SDS-PAGE becomes severely compromised when more than 100-150 µg protein is applied. The microsomal fraction used by Kraut and associates (21) consists, most likely, of a mixture of cellular membranes of different origin (plasma membranes, Golgi membranes, endoplasmic reticulum, etc). In our experiments we used purified plasma membranes, because we reasoned that this fraction would provide a better approximation of the location of the functional H⁺-K⁺-ATPase. The failure to demonstrate HK₁ or HK₂ consistently in control conditions by typical Western blot analysis indicates a low level of expression of these proteins in whole kidney. Since the findings of Kraut et al. (21) for whole kidney microsomes also provide evidence that HK₁, unlike HK₂, does not respond to chronic hypokalemia, the findings of these investigators appear to be in general agreement with our results and support the view that HK₁ is not upregulated significantly by chronic hypokalemia. However, because of the low level of HK₁ in kidney and our inability to detect HK₁ expression in control rats, we cannot rule out the possibility that hypokalemia has a potential effect on HK₁ abundance below the level of detection of our method. Nevertheless, we have been able to localize this regulatory response for HK₂ protein to the renal medulla.

As in whole kidney, previous studies from our laboratory reported a two- to threefold increase in HK₂ mRNA in whole distal colon following chronic dietary K⁺ depletion (9). Although the distal colon represents another important site for potassium absorption during chronic hypokalemia (7, 9, 29, 32), we were unable, in the present study, to verify a similar regulatory response for HK₂ protein in distal colon under identical experimental conditions. The explanation for the lack of correlation between mRNA and protein abundance in distal colon during chronic hypokalemia is not entirely clear. It is conceivable that hypokalemia could produce changes in translational efficiency in the distal colon, so that more mRNA would be required to synthesize the same amount of protein. Alternatively, there could be an increase in protein turnover, so that the synthetic rate would need to be enhanced to maintain protein levels. Since the increase in HK₂ mRNA abundance in distal colon after chronic hypokalemia has not been observed by other laboratories (18, 29), it is also possible that our report was incorrect. Our observation that the increase in HK₂ mRNA in distal colon was consistent and highly reproducible (9) suggests that an error in technique is highly unlikely, nevertheless.

Our results also indicate that the level of expression of ₁-Na⁺-K⁺-ATPase did not change in either the renal cortex or the renal medulla with chronic dietary K⁺ depletion or CMA. Barlet-Bas et al. (5), Hayashi and Katz (16), and McDonough et al. (25) reported that ₁-Na⁺-K⁺-ATPase is expressed abundantly in all regions of the nephron with the exception of the proximal straight tubule and MCD. The ₁-Na⁺-K⁺-ATPase in the MCD represents ~15% of the total ₁-Na⁺-K⁺-ATPase in the renal medulla. The abundance of ₁-Na⁺K⁺-ATPase in the MCD is increased twofold with chronic dietary K⁺ depletion (25). This increase represents ~25% of the total ₁-Na⁺-K⁺-ATPase present in renal medulla (25). An increase of this magnitude would not be detected by the immunoblot system employed in the present study, however. Nevertheless, since ₁-Na⁺-K⁺-ATPase abundance in whole kidney and in kidney regions greatly exceeds the abundance of H⁺-K⁺-ATPase, our purpose in measuring ₁-Na⁺-K⁺-ATPase was to provide an internal standard. Since ₁-Na⁺-K⁺-ATPase and HK₂ did not increase in parallel with hypokalemia, it is likely that changes in HK₂ abundance with hypokalemia were not the result of a nonspecific effect on cell size (hypertrophy).

Although we found no effect of CMA on expression of HK₁ or HK₂ protein in either renal cortex or medulla, Silver et al. (31) have demonstrated clearly that CMA increased the rate of Sch-28080-sensitive pH_i recovery after an acid load. The model of CMA employed by these investigators was significantly different from that employed in this study, since NH₄Cl ingestion was over 10-14 days rather than 4 days. Sch-28080 sensitivity was employed as the "marker" of H⁺-K⁺-ATPase activity in that study (31) so that it is impossible to determine which specific H⁺-K⁺-ATPase isoform is affected by CMA of this duration. Further studies on HK₁ and HK₂ protein abundance during prolonged NH₄Cl ingestion are indicated, therefore, and will be the subject of future studies in our laboratory.

In conclusion, our data demonstrate that HK₂ protein is expressed in the renal medulla and distal colon. The abundance of HK₂ protein was augmented by chronic hypokalemia significantly in renal medulla, not in distal colon. CMA did not alter the abundance of HK₂ protein in either renal medulla or distal colon. These findings substantiate a parallel regulatory response by chronic hypokalemia for HK₂ mRNA and protein which can be localized to the renal medulla.