Renal ammonia excretion in response to hypokalemia: effect of collecting duct-specific Rh C glycoprotein deletion

Abstract

The Rhesus factor protein, Rh C glycoprotein (Rhcg), is an ammonia transporter whose expression in the collecting duct is necessary for normal ammonia excretion both in basal conditions and in response to metabolic acidosis. Hypokalemia is a common clinical condition associated with increased renal ammonia excretion. In contrast to basal conditions and metabolic acidosis, increased ammonia excretion during hypokalemia can lead to an acid-base disorder, metabolic alkalosis, rather than maintenance of acid-base homeostasis. The purpose of the current studies was to determine Rhcg's role in hypokalemia-stimulated renal ammonia excretion through the use of mice with collecting duct-specific Rhcg deletion (CD-Rhcg-KO). In mice with intact Rhcg expression, a K⁺-free diet increased urinary ammonia excretion and urine alkalinization and concurrently increased Rhcg expression in the collecting duct in the outer medulla. Immunohistochemistry and immunogold electron microscopy showed hypokalemia increased both apical and basolateral Rhcg expression. In CD-Rhcg-KO, a K⁺-free diet increased urinary ammonia excretion and caused urine alkalinization, and the magnitude of these changes did not differ from mice with intact Rhcg expression. In mice on a K⁺-free diet, CD-Rhcg-KO increased phosphate-dependent glutaminase (PDG) expression in the outer medulla. We conclude that hypokalemia increases collecting duct Rhcg expression, that this likely contributes to the hypokalemia-stimulated increase in urinary ammonia excretion, and that adaptive increases in PDG expression can compensate for the absence of collecting duct Rhcg.

renal ammonia metabolism is a primary method through which the kidneys maintain acid-base homeostasis, and involves integrated contributions of both intrarenal ammoniagenesis and renal epithelial cell ammonia transport.¹ Increasingly, the “traditional” theory of passive NH₃ diffusion and NH₄⁺ trapping is being replaced by one in which specific membrane proteins transport NH₃ or NH₄⁺, enabling facilitated and regulated membrane NH₃ and NH₄⁺ transport (32, 33). Prominent among these ammonia-transporting proteins are the Rh glycoproteins. Three mammalian Rh glycoproteins have been identified, Rh A glycoprotein, which is expressed specifically in erythroid cells, and the nonerythroid Rh glycoproteins, Rh B glycoprotein (Rhbg) and Rh C glycoprotein (Rhcg), which are expressed widely in ammonia-transporting tissues, including the kidney (29, 30, 32, 33).

In the kidney, Rhcg is expressed in the renal distal nephron and collecting duct, from the distal convoluted tubule (DCT) through the inner medullary collecting duct (8, 10, 20, 28), where 60–80% of final urinary ammonia is secreted. Evidence supporting a critical role for Rhcg in renal ammonia transport includes findings that 1) Rhcg transports NH₃ (reviewed in Ref. 32); 2) changes in Rhcg expression parallel ammonia excretion in multiple conditions, including metabolic acidosis and reduced renal mass (15, 17, 21, 22); 3) Rhcg gene deletion impairs both basal and acidosis-stimulated renal ammonia excretion (3, 17, 18); and 4) Rhcg deletion decreases apical NH₃ permeability in isolated, perfused collecting ducts (3).

Hypokalemia is a common clinical condition and typically gives rise to increased renal ammonia excretion (23, 26). However, in contrast to both metabolic acidosis and reduced renal mass, conditions in which increased single-nephron ammonia excretion facilitates normal acid-base homeostasis, the increased renal ammonia excretion in hypokalemia is not adaptive in terms of acid-base homeostasis; in many species, including both rats and humans, hypokalemia can cause development of an acid-base disorder, namely, metabolic alkalosis (6, 11, 13). We reported recently that hypokalemia induced by dietary K⁺ deficiency in the rat increased urinary ammonia excretion, caused metabolic alkalosis, and increased Rhcg expression in the outer medulla in both intercalated cells and principal cells (11). These results suggest that Rhcg expression in the renal collecting duct may mediate an important role in the increased urinary ammonia excretion in response to hypokalemia.

The purpose of this study was to examine the effect of collecting duct Rhcg deletion on the renal response to hypokalemia. First, we determined the effect of hypokalemia, generated by dietary K⁺ deficiency, on mice with intact Rhcg expression. Next, we examined whether collecting duct-specific Rhcg deletion (CD-Rhcg-KO), generated using loxP-Cre techniques we described previously (17, 18), alters renal responses to hypokalemia. Finally, we examined adaptive responses to collecting duct-specific Rhcg deletion in hypokalemic mice that could compensate for the lack of Rhcg-mediated NH₃ transport. Our results show that Rhcg expression increases in the mouse outer medullary collecting duct (OMCD) in response to dietary K⁺ deficiency. Collecting duct deletion of Rhcg in hypokalemic mice did not alter either urinary ammonia excretion or urine pH. However, collecting duct Rhcg deletion from hypokalemic mice did increase renal expression of phosphate-dependent glutaminase (PDG), the major enzyme involved in renal ammoniagenesis. These results indicate that increased collecting duct Rhcg expression is a component of the renal response to hypokalemia and likely contributes to the increased ammonia excretion but that adaptive responses to Rhcg deletion during hypokalemia can compensate for the absence of collecting duct Rhcg-mediated NH₃ transport and enable normal increases in renal ammonia excretion.

METHODS

Animals.

Mice used in this project were generated by mating mice homozygous for floxed Rhcg alleles and expressing Cre-recombinase under control of the Ksp-cadherin promoter (Ksp-cadherin-Cre) with mice homozygous for floxed Rhcg alleles but not expressing Ksp-cadherin-Cre, as described previously (17). Trained personnel at the University of Florida College of Medicine Cancer and Genetics Transgenic Animal Core Facility performed all animal breeding. Mice were genotyped using DNA obtained from tail-clip samples as we described previously (17, 18). We have shown previously that mice homozygous for floxed Rhcg alleles that express Ksp-cadherin-Cre exhibit collecting duct-specific Rhcg deletion whereas control mice, homozygous for floxed Rhcg alleles but not expressing Ksp-cadherin-Cre, have intact Rhcg expression in a pattern not detectably different from that observed in wild-type mice (17). All studies used both male and female mice, and responses to experimental conditions did not differ in different-sex mice. All animal studies were approved by the University of Florida College of Medicine and the North Florida/South Georgia Veterans Health System Institutional Animal Care and Use Committees.

Antibodies.

Affinity-purified antibodies to Rhbg and Rhcg generated in our laboratory have been characterized previously (19, 28, 31). Antibodies to PDG were provided by Dr. Norman Curthoys (Colorado State University), and antibodies to the a4 subunit of H⁺-ATPase were provided by Dr. Fiona Karet (Cambridge Institute for Medical Research, Cambridge, UK). Antibodies to phosphoenolpyruvate carboxykinase (PEPCK) were obtained from Cayman Chemical (Ann Arbor, MI), and antibodies to glutamine synthetase (GS) were obtained from Chemicon (Temecula, CA).

Artificial diets.

Powdered semisynthetic diets were used to enable altered dietary cation content. These include a K⁺-free diet and K⁺ control [identical to K⁺ free but with normal K⁺ content (TD.88239, TD.88238, respectively) obtained from Harlan Teklad, Madison, WI]. Similar results were obtained with a K⁺ control diet and a regular diet (TD 2018, powdered formulation), and results were combined for analysis. Diet with 25% of normal K⁺ content was prepared by mixing the K⁺-free diet with normal rodent chow in a ratio of 3:1. All diets were obtained in a powdered state and mixed with water to generate semisolid food.

Metabolic cage studies.

Animals were placed in metabolic cages (Tecniplast diuresis metabolic cage, Fisher Scientific) and allowed to acclimate for 1 day before measurements were begun. Mice were allowed ad libitum access to water and were given a preweighed amount of fresh food each day; we calculated daily food intake by subtracting the amount remaining after 24 h. All urine was collected under water-equilibrated mineral oil, and daily urine volume and pH were recorded. Urine samples were then stored at −20°C until analyzed further.

Electrolyte measurements.

Urine ammonia was measured using a modification of a commercially available kit (A7553, Pointe Scientific, Canton, MI) as described previously (17). Urine pH was measured using a micro-pH electrode (ROSS semi-micro pH, Orion 8115BN, Thermo Scientific). Serum bicarbonate was measured as total CO₂ using a commercially available kit (C750-120, Pointe Scientific) as described previously (17). Plasma and urine K⁺ were measured using a flame photometer (Instrumentation Laboratory, Lexington, MA). Plasma and urine creatinine were measured using capillary electrophoresis as described previously (34), with the exception that the injection time was 10 s, rinses used 0.1 M NaOH, LC/MS grade water, and then running buffer, with 2-min rinses between runs, and the detection wavelength was 214 nm.

Tissue preparation for immunolabeling.

Mice were anesthetized with inhalant isoflurane. Kidneys were preserved by in vivo cardiac perfusion with PBS (pH 7.4) followed by periodate-lysine-2% paraformaldehyde (PLP), then cut transversely into several 2- to 4-mm thick slices and immersed for 24–48 h at 4°C in the same fixative. Kidney samples from each animal were embedded in polyester wax made using polyethylene glycol 400 distearate (Polysciences, Warrington, PA) with 10% 1-hexadecanol, and 2- or 3-μm-thick sections were cut and mounted on gelatin-coated glass slides.

For immunogold cytochemistry, ∼1-mm³ samples of outer stripe of outer medulla were treated with 0.1 M NH₄Cl for 1 h, dehydrated in a graded series of ethanols, and infiltrated and embedded in Lowicryl K4M (Polysciences). Blocks were polymerized under UV light for 24 h at −20°C, followed by ∼60 h at room temperature; then, 0.5-μm sections were stained with toluidine blue, and samples containing well-preserved outer stripe of the OMCD (OMCDo) were selected for electron microscopy. Ultrathin sections, ∼60-nm thick, were mounted on formvar/carbon-coated nickel grids for immunogold labeling.

Immunohistochemistry.

Immunolocalization was accomplished using immunoperoxidase procedures described previously (14, 15, 17). Briefly, sections were dewaxed in ethanol, rehydrated, and then rinsed in PBS. Endogenous peroxidase activity was blocked by incubating the sections in 3% H₂O₂ in distilled water for 45 min. The sections were blocked for 15 min with Serum-Free Protein Block (DakoCytomation) and were then incubated at 4°C overnight with primary antibody. The sections were washed in PBS and incubated for 30 min with polymer-linked, peroxidase-conjugated goat anti-rabbit IgG (MACH2, Biocare Medical, Concord, CA), again washed with PBS, and then exposed to diaminobenzidine (DAB) for 5 min. The sections were washed in distilled water, dehydrated with xylene, mounted, and observed by light microscopy. Comparisons of labeling were made only between sections of the same thickness from the same immunohistochemistry experiment. Sections were examined on a Nikon E600 microscope equipped with DIC optics and photographed using a DXM1200F digital camera and ACT-1 software (Nikon). Color correction was performed using Adobe Photoshop software (Adobe Systems, San Jose, CA).

Protein preparation for immunoblot analysis.

Animals were anesthetized with inhalant isoflurane, and the kidneys were rinsed by in vivo cardiac perfusion with PBS (pH 7.4). The right renal vasculature was clamped and the right kidney was rapidly removed, the cortex and outer medulla were isolated rapidly under a dissecting microscopy, snap-frozen in liquid nitrogen, and then stored frozen at −70°C until used. The left kidney was perfused with PLP fixative for immunohistochemistry. Tissues were homogenized in Tissue Protein Extraction Reagent (T-PER; Pierce Biotechnology, Rockford, IL) using microtube pestles (USA Scientific, Ocala, FL), and protein was extracted according to the manufacturer's recommended procedures. An aliquot was obtained for protein determination using a BCA assay, and the remainder was stored frozen at −70°C until used. For membrane protein preparation for Rhcg, tissues were homogenized in buffer A (in mM: 50 sucrose, 10 Tris buffer, 1 EDTA, pH 7.4) and then diluted in buffer B (in mM: 250 sucrose, 10 Tris buffer, and 1 EDTA, pH 7.4). The sample was then centrifuged at 1,000 g for 5 min at 4°C. The pellet was resuspended in buffer B and again centrifuged at 21,000 g for 30 min at 4°C. The 21,000-g pellet was finally resuspended in buffer B. An aliquot was used for protein determination with a BCA assay, and the remainder was stored frozen at −80°C until used.

Immunogold labeling.

Briefly, the immunogold labeling procedure was performed by exposure of the ultrathin tissue sections to the primary antibody and then to a goat anti-rabbit IgG secondary antibody conjugated to 10-nm colloidal gold particles (British BioCell International, Ted Pella, Redding, CA). Unless noted otherwise, all steps were done by floating the grids on droplets of solution at room temperature. The sections were exposed to 0.1 M NH₄Cl for 1 h, rinsed with Rinse buffer (0.01 M Tris·HCl and 0.15 M NaCl, in distilled water, pH 7.2), treated with BSA buffer (1% BSA, 0.01 M Tris·HCl, 0.5 M NaCl, and 4 mM NaN₃, in distilled water, pH 7.2) for 30 min, and then incubated in a humidified chamber overnight at 4°C with the Rhcg primary antibody diluted in filtered BSA buffer. The sections were washed with Rinse buffer, treated with Carbowax buffer (0.02% Carbowax PEG 20M, 0.01 M Tris·HCl, 0.15 M NaCl, and 4 mM NaN₃, in distilled water, pH 7.2) for 30 min, and then exposed to the secondary antibody diluted in Carbowax buffer for 1 h. The sections were washed with Rinse buffer, washed with PBS, postfixed with 1.6% glutaraldehyde in PBS, washed with PBS, and then with distilled water. Silver enhancement of the gold particles was done using Aurion EM SE (Electron Microscopy Sciences, Hatfield, PA) for 45 min, and then the grids were washed again with distilled water, dried overnight, and counterstained with saturated uranyl acetate. Each group of sections subjected to the immunogold procedure included a control section that was exposed to BSA buffer in place of the primary antibody. Ultrathin sections were examined using a Hitachi 7600 transmission electron microscope (Hitachi High-Technologies America, Pleasanton, CA) equipped with a Macrofire monochrome progressive scan charge-coupled device camera (Optronics, Goleta, CA) and AMT image-capture software (Advanced Microscopy Techniques, Danvers, MA). The OMCDo was identified by its characteristic heterogeneous epithelial cell population, which included principal cells and intercalated cells.

Morphometric analysis.

Gold label along the apical plasma membrane was quantified in type A intercalated cells in the OMCD in four control and five hypokalemic mouse kidneys. In each animal, at least four cells of type A intercalated cell were selected randomly and photographed. The number of gold particles along the apical plasma membranes was counted manually. The mean apical plasma membrane gold particle count/cell was calculated for each animal, and these values were used for the statistical analysis.

Immunoblotting procedure.

Ten micrograms of renal protein were electrophoresed on 10% PAGE ReadyGel (Bio-Rad, Hercules, CA). Gels were then transferred electrophoretically to nitrocellulose membranes, blocked with 5 g/dl nonfat dry milk, and incubated at 4°C overnight with primary antibody diluted in Blotto buffer (50 mM Tris, 150 mM NaCl, 5 mM Na₂EDTA, and 0.05% Tween 20, pH 7.6) with 5 g/dl nonfat dry milk. Loading and transfer equivalence were assessed with Ponceau S staining. After washing, membranes were exposed to a secondary antibody, goat anti-rabbit IgG (Millipore, Billerica, MA) or goat anti-mouse IgG (Upstate, Temecula, CA), conjugated to horseradish peroxidase at a dilution of 1:5,000. Sites of antibody-antigen reaction were visualized by using enhanced chemiluminescence (SuperSignal West Pico Substrate, Pierce) and a Kodak Image Station 440CF digital imaging system. Band density was quantified using Kodak 1D, version 5.0, software (Kodak Scientific Imaging, New Haven, CT). Band density was normalized such that mean density in the same region (cortex or outer medulla) in control tissues was 100.0. Absence of saturation was confirmed by examining pixel intensity distribution in all immunoblots.

Statistics.

Results are presented as means ± SE. Statistical analyses were performed using Student's t-test, and P < 0.05 was taken as statistically significant; n refers to the numbers of animal studied.

RESULTS

Physiological effect of dietary K⁺ deficiency in mice.

In the first experiments, we determined the effect of dietary K⁺ deficiency for 12 days on serum electrolytes in the mouse. Serum K⁺ concentration decreased significantly, from 4.00 ± 0.20 to 3.12 ± 0.20 mmol/l (P < 0.02, n = 4 in each group). Serum Na⁺ did not change significantly (data not shown). Thus 12 days of a K⁺-free diet caused hypokalemia.

We next determined the effect of hypokalemia induced by dietary K⁺ deficiency on urinary parameters. Urine total ammonia excretion increased rapidly, as shown in Fig. 1A. Simultaneously, urinary K⁺ excretion decreased, with an ∼80% decrease on the first day and a 95% decrease by day 2 of the K⁺-free diet (Fig. 1B). Urine pH increased in mice on the K⁺-free diet from baseline of 5.97 ± 0.05 to 7.07 ± 0.13 on day 2 of hypokalemia and 7.28 ± 0.05 on day 12 of hypokalemia (P < 0.001 for each comparison, n = 8).

Fig. 1.

Urinary ammonia excretion and urinary K⁺ excretion in response to a K⁺-free diet for 12 days in control mice. Mice were housed in metabolic cages, and urine was collected daily under mineral oil. A: urinary ammonia excretion. A K⁺-free diet resulted in significant increases in urinary ammonia excretion. *P < 0.02 vs. a normal-K⁺ diet. B: urinary K⁺ excretion. A K⁺-free diet resulted in rapid decreases in urinary potassium excretion. *P < 0.05 vs. K⁺ excretion while on a normal-K⁺ diet; n = 8/group.

Rhcg expression in response to hypokalemia.

Recent studies in the rat show that a K⁺-free diet increases Rhcg expression in the outer medulla, but not the cortex (11). Similarly, in the mouse, a K⁺-free diet increased Rhcg protein expression significantly in the outer medulla, but not the cortex (Fig. 2, A and B) [cortex: K⁺-free, n = 8, 113.7 ± 4.7 vs. K⁺ control, n = 4, 100.0 ± 6.5; P = not significant (NS); outer medulla: K⁺-free, n = 8, 407.4 ± 13.8 vs. K⁺ control, n = 4, 100.0 ± 8.2; P < 0.000001].

Fig. 2.

Rh C glycoprotein (Rhcg) expression in control mice fed a K⁺ control and K⁺-free diet. Shown are an immunoblot analysis (A) and quantification of Rhcg expression in the cortex and outer medulla (B). Feeding mice a K⁺-free diet for 12 days increased Rhcg protein expression significantly in the outer medulla, but not in the cortex. Expression normalized to mean expression was equal to 100 for the K⁺ control diet in each region for each protein. Values are means ± SE; n = 4/group on a K⁺ control diet and n = 8/group on a K⁺-free diet. NS, not significant. C: Rhcg immunolabel in the outer medulla of mice on a K⁺ control and K⁺-free diet for 12 days. The K⁺-free diet induced hypertrophy of both intercalated cells and principal cells, and it increased Rhcg expression in the outer medullary collecting duct (OMCD). High magnifications show that basolateral Rhcg immunolabel intensity was increased in the OMCD by a K⁺-free diet.

Immunohistochemistry demonstrated dramatic changes in Rhcg immunolabel expression in the OMCD (Fig. 2C). In the OMCD, both apical and basolateral Rhcg immunolabel intensity was substantially increased in response to hypokalemia in both intercalated and principal cells. Thus, similar to in the rat kidney (11), hypokalemia increased both apical and basolateral Rhcg immunolabel.

To confirm that apical Rhcg immunolabel intensity was increased, we used immunogold electron microscopy with quantitative morphometric analysis to assess apical plasma membrane Rhcg expression. Figure 3 shows representative micrographs from control and hypokalemia kidneys, and Fig. 3C shows the quantitative analysis. Hypokalemia for 12 days increased apical Rhcg immunolabel expression significantly in intercalated cells in the outer strip of the outer medulla (control diet, 161 ± 14; K⁺-free diet, 226 ± 20 gold particles/cell, P < 0.02, n = 4 and 5, respectively). These observations confirm the light microscopic immunohistochemistry findings that hypokalemia increases apical Rhcg immunolabel.

Fig. 3.

Immunogold electron microscopy for Rhcg and effect of hypokalemia. A and B: representative images from control (A) and hypokalemic (B) kidneys of apical region of intercalated cells in the outer stripe of the outer medulla labeled for Rhcg. Apical plasma membrane Rhcg (arrows) was increased in the hypokalemic kidney. C: quantitative analysis of apical plasma membrane Rhcg immunolabel. Apical Rhcg immunolabel was increased significantly in hypokalemic mice.

We next performed triple-immunolabel studies, using serial kidney sections, with antibodies to Rhcg in one section and double-immunolabel staining in the second section with antibodies to H⁺-ATPase and Rhbg to enable assessment of Rhcg expression separately in intercalated and principal cells. These studies confirmed that in the hypokalemic kidney basolateral intensity was similar in intercalated cells (identified by apical H⁺-ATPase and basolateral Rhbg immunolabel) and principal cells (identified by the absence of apical H⁺-ATPase and Rhbg) (Fig. 4). There was no detectable change in Rhcg immunolabel expression in the cortex in response to hypokalemia.

Fig. 4.

Colocalization in the inner stripe of the OMCD of Rhcg with H⁺-ATPase and Rh B glycoprotein (Rhbg) in mice on a control and K⁺-free diet for 12 days. A: Rhcg immunolabel in the inner stripe of the OMCD of mouse on a control diet. B: serial section with double-immunolabel for Rhbg (brown) and H⁺-ATPase (blue). Intercalated cells (arrows) are identified by apical H⁺-ATPase and basolateral Rhbg immunolabel (arrows in both A and B) and principal cells by the absence of H⁺-ATPase and Rhbg immunolabel (arrowheads). Intercalated cells have stronger apical Rhcg immunolabel than do principal cells. C and D: similar Rhcg immunolabel in mice on a K⁺-free diet for 12 days. D: serial section of C double-immunolabeled for H⁺-ATPase (blue) and Rhbg (brown). Again, intercalated cells (arrows) are identified by apical H⁺-ATPase and basolateral Rhbg immunolabel, and principal cells (arrowheads) by the absence of these markers. Increased Rhcg immunolabel intensity is evident in mice on a K⁺-free diet, with increases in basolateral Rhcg immunolabel intensity in both intercalated cells and principal cells. Apical Rhcg immunolabel intensity was not reliably increased in either intercalated cells or principal cells. Results are representative of studies performed in 4 mice on a control diet and 4 on a K⁺-free diet.

Effect of CD-Rhcg-KO on response to a K⁺-free diet.

The increased Rhcg expression observed in response to hypokalemia in both the mouse (current studies) and the rat (11) suggests that increased Rhcg expression contributes to the observed increases in renal ammonia excretion. To test this, we determined whether Rhcg expression was required for the hypokalemia-induced increase in urinary ammonia excretion. We used mice with targeted, collecting duct-specific Rhcg deletion to avoid unexpected effects of Rhcg deletion related to unexpected roles of Rhcg in other tissues in which it is expressed. Thus we compared mice with either intact or collecting duct-specific Rhcg deletion fed a K⁺-free diet. Food intake, and thus K⁺ intake, was measured daily and did not differ significantly between control and CD-Rhcg-KO mice (data not shown).

Results of plasma electrolyte analysis are summarized in Table 1. Mice with intact Rhcg expression and with CD-Rhcg-KO developed significant hypokalemia in response to 12 days of a K⁺-free diet. Serum K⁺ concentration, however, did not differ significantly between the two groups of mice. Serum HCO₃⁻ did not differ significantly between mice with intact Rhcg expression and those with CD-Rhcg-KO. Although endogenous creatinine clearance overestimates actual glomerular filtration rate (GFR) due to organic anion transporter 3-mediated creatinine secretion (7, 27), we used serum creatinine and creatinine clearance as estimates of the glomerular filtration rate (GFR) and found that collecting duct-specific Rhcg deletion did not alter creatinine clearance significantly in hypokalemic mice. Thus deletion of Rhcg from intercalated cells does not alter serum K⁺, serum HCO₃⁻, or apparent GFR in mice provided a K⁺-free diet for 12 days.

Table 1.

Physiological parameters after K⁺-free diet for 12 days

Parameter	Control (n = 4)	CD-Rhcg-KO (n = 4)	P Value
Serum HCO3−, mmol/l	19.0 ± 1.5	19.1 ± 0.9	NS
Serum K+, mmol/l	3.2 ± 0.2	3.3 ± 0.1	NS
Serum creatinine, mg/dl	0.061 ± 0.002	0.056 ± 0.007	NS
Creatinine clearance, μl/min	421 ± 11	412 ± 75	NS

Values are means ± SE. CD-Rhcg-KO, collecting duct-specific Rh C glycoprotein (Rhcg) deletion; NS, not significant.

Next, we determined the effect of collecting duct-specific Rhcg deletion on the expected changes in urinary ammonia, pH, and K⁺ excretion in response to hypokalemia. To ensure that we did not miss any time-dependent differences, we obtained 24-h urine collections both before the K⁺-free diet and on each day of the 12 days of the K⁺-free diet. Both control and CD-Rhcg-KO mice responded to the K⁺-free diet with an increase in urinary ammonia excretion; changes were evident on the first day of the K⁺-free diet and increased until a maximum on days 3–4, after which ammonia excretion remained constant (Fig. 5A). On no day, with the exception of day 6, did urinary ammonia excretion differ significantly between mice with intact Rhcg expression and those with CD-Rhcg-KO.

Fig. 5.

Effect of collecting duct-specific Rhcg deletion (CD-Rhcg-KO) on response to hypokalemia. Mice were housed in metabolic cages, and urine was collected daily. A: urine ammonia excretion. A K⁺-free diet increased urine ammonia excretion in mice with CD-Rhcg-KO and with intact Rhcg expression. There was no significant difference in urinary ammonia excretion in mice with CD-Rhcg-KO, with the exception of day 6; n = 12–20 in mice with intact Rhcg expression and 8–15 in CD-Rhcg-KO mice. B: urine pH. A K⁺-free diet resulted in urinary alkalinization, with significant increases in urine pH evident on the first day. There was no difference in urine pH on any day between mice with CD-Rhcg-KO and mice with intact Rhcg expression; n = 12–20 on each day. C: urinary K⁺ excretion. Open diamonds and dotted lines are used to indicate K⁺ excretion by mice with intact Rhcg expression, and open squares with solid lines indicate K⁺ excretion by CD-Rhcg-KO mice. When SE bars are not shown, they are smaller than the size of the marker. Urinary K⁺ excretion decreased rapidly with a K⁺-free diet, and there were no differences in K⁺ excretion between mice with intact Rhcg expression and CD-Rhcg-KO mice.

Provision of the K⁺-free diet had significant effects on urine pH, causing substantial urine alkalinization (Fig. 5B). Initial changes in urine pH occurred more rapidly than changes in urine ammonia excretion, by approximately 1 day. Similar to the findings with regard to ammonia excretion, urine pH did not differ significantly between mice with intact Rhcg expression and those with CD-Rhcg-KO.

The finding that hypokalemia results in concomitant increases in urinary ammonia excretion and decreases in K⁺ excretion has sometimes suggested that these changes are causally interrelated. To begin examining this possibility and to ensure that differences in renal K⁺ excretion as a consequence of Rhcg deletion did not induce greater degrees of K⁺ depletion, we examined urinary K⁺ excretion (Fig. 5C). Significant decreases in urinary K⁺ excretion (∼65% decrease) were evident on the first day of the K⁺-free diet and plateaued with an ∼95–98% decrease. Urinary K⁺ excretion did not differ between mice with intact Rhcg expression and those with CD-Rhcg-KO significantly, either before the K⁺-free diet or on any day of the 12 days of the K⁺-free diet. Thus collecting duct-specific Rhcg deletion does not alter renal K⁺ excretion in response to a K⁺-free diet.

Effect of 75% reduction in dietary K⁺.

To examine Rhcg's role in the response to more modest degrees of K⁺ deficiency, we examined the effect of collecting duct-specific Rhcg deletion on the response to a diet with a 75% reduction in dietary K⁺, i.e., 25% of the usual K⁺ content. Neither serum HCO₃⁻ nor serum K⁺ concentrations differed significantly between mice with intact and those with CD-Rhcg-KO (Table 2). Thus deletion of Rhcg from the renal collecting duct does not alter significantly either serum K⁺ or HCO₃⁻ in response to a 75% reduction in dietary K⁺ availability for 3 days.

Table 2.

Physiological parameters after 75% K⁺ restriction for 3 days

	Control	CD-Rhcg-KO	P Value
Serum HCO3−, mmol/l	20.7 ± 2.0 (4)	20.6 ± 0.7 (4)	NS
Serum K+, mmol/l	4.3 ± 0.2 (4)	4.5 ± 0.2 (4)	NS

Values are means ± SE. Numbers in parentheses are numbers of animals in each group.

Urine ammonia excretion increased on days 1, 2, and 3 (P < 0.05 by paired t-test), but there was no difference in total urinary ammonia between mice with intact Rhcg expression and those with CD-Rhcg-KO (P = NS, n = 4/group). Figure 6A summarizes these results. Urine pH increased significantly on day 2 (P < 0.05), and there was a trend for an increase on days 1 and 3 that did not reach statistical significance (P = NS). Again, there was no statistically significant difference between the urine pH of mice with intact Rhcg expression and those with CD-Rhcg-KO (P = NS, n = 4/group). Figure 6B shows these findings. Thus deletion of Rhcg from the collecting duct does not alter significantly either urinary ammonia excretion or pH in response to a 75% decrease in dietary K⁺.

Fig. 6.

Urinary ammonia excretion, urine pH, and urinary K⁺ excretion in response to 75% decrease in dietary K⁺ content. A diet with only 25% of the K⁺ content of normal was prepared by mixing powdered formulations of the K⁺-free and K⁺-containing diet in a ratio of 3:1. A: urinary ammonia excretion. Partial dietary K⁺ restriction increased urinary ammonia excretion on each day [*P < 0.05 vs. before the diet (Pre)], but urinary ammonia excretion did not differ significantly between mice with intact Rhcg expression and those with CD-Rhcg-KO. B: urine pH in response to decreased dietary K⁺ content. A 75% reduction in dietary K⁺ content resulted in increased urine pH on day 2 (*P < 0.05 vs., Pre), and there were trends for increases on days 1 and 3. There was no difference in urine pH between mice with intact Rhcg expression and those with CD-Rhcg-KO on any day; n = 4. C: urinary K⁺ excretion. Restricted dietary K⁺ intake resulted in significant reductions in urinary K⁺ excretion (*P < 0.05 vs. Pre), but there was no difference in urinary K⁺ excretion, either before the K⁺-free diet or on any day of the restricted K⁺ diet, between mice with intact Rhcg expression and those with CD-Rhcg-KO (P = NS; n = 4 in each group). SE bars, when not visible, are smaller than the size of the data symbol.

Finally, urinary K⁺ excretion during each day of restricted K⁺ intake was significantly less in animals than while on their usual dietary K⁺ intake (P < 0.05 vs. before the K⁺-free diet). However, K⁺ excretion by mice with intact Rhcg expression and mice with CD-Rhcg-KO did not differ significantly (P = NS, n = 4/group). Figure 6C summarizes these results. Thus collecting duct-specific Rhcg deletion does not alter either urinary ammonia excretion, pH, or K⁺ excretion in response to a 75% dietary K⁺ restriction.

Effect of a K⁺-free diet on proteins involved in ammonia metabolism.

Because deletion of collecting duct Rhcg did not affect renal ammonia excretion during hypokalemia, we hypothesized that other mechanisms involved in renal ammonia metabolism and excretion might be upregulated in hypokalemic CD-Rhcg-KO mice. The next set of studies was designed to examine whether increased expression of enzymes involved in ammonia production and metabolism compensated for the absence of collecting duct Rhcg expression. PDG and PEPCK are the primary enzymes involved in renal ammoniagenesis, and GS mediates an important role in net ammoniagenesis because of its ability to catalyze reaction of NH₄⁺ with glutamate to form glutamine, thereby decreasing net ammoniagenesis.

First, we determined the effect of hypokalemia induced by dietary K⁺ deficiency on PEPCK, PDG, and GS expression in the mouse. Figure 7 summarizes these findings. Hypokalemia increased expression of both PEPCK and PDG significantly in the renal cortex, but not in the outer medulla. GS expression decreased significantly in both the cortex and the outer medulla. Thus this coordinated response, increased expression of enzymes involved in ammonia production and decreased expression of the enzyme GS, which catalyzes decreased net ammonia catabolism, results in synergistic effects to increase net renal ammoniagenesis.

Fig. 7.

Effect of hypokalemia on expression of enzymes involved in ammoniagenesis. A: effect of hypokalemia on phosphate-dependent glutaminase (PDG) expression. Hypokalemia increased PDG expression significantly in the cortex. There was a slight decrease in expression in the outer medulla (OM). B: immunoblot analysis of phosphoenolpyruvate carboxykinase (PEPCK) protein expression. Hypokalemia increased PEPCK expression in the cortex significantly. There is no significant change in PEPCK expression in the OM. C: effects of hypokalemia on glutamine synthetase (GS) expression. Hypokalemia resulted in dramatic and significant decreases in GS expression in both the cortex and the OM.

Next, we examined whether the magnitude of these increases differed between mice with intact and with collecting duct-specific Rhcg deletion treated with a K⁺-free diet. Figure 8 summarizes these results. When examined in hypokalemic mice, PDG, the primary enzyme involved in renal ammoniagenesis, was expressed at significantly higher levels in the outer medulla of CD-Rhcg-KO mice than in mice with intact Rhcg expression. However, CD-Rhcg-KO did not significantly alter either PEPCK or GS expression in hypokalemic mice. In addition, Rhbg expression, which also increases in response to hypokalemia (1), did not differ between hypokalemic mice with intact Rhcg expression and those with CD-Rhcg-KO.

Fig. 8.

Effect of CD-Rhcg-KO on expression of ammoniagenic enzymes in hypokalemic mice. A: PDG expression in hypokalemic mice with intact Rhcg expression (“control”) and with CD-Rhcg-KO. In hypokalemic mice, CD-Rhcg-KO did significantly increase PDG expression in the outer medulla, but not in the cortex. B: PEPCK expression in hypokalemic mice with intact Rhcg expression and hypokalemic mice with CD-Rhcg-KO. There was no significant difference in PEPCK expression in either the cortex or the outer medulla in hypokalemic mice as a consequence of CD-Rhcg-KO. C: GS expression in the cortex and outer medulla of hypokalemia mice with intact Rhcg expression and with CD-Rhcg-KO. In hypokalemic mice, GS expression did not differ significantly between mice with intact Rhcg expression and those with CD-Rhcg-KO. D: Rhbg expression in the cortex and outer medulla. In hypokalemic mice, CD-Rhcg-KO did not result in adaptive changes in Rhbg expression in either the cortex or the outer medulla. Expression normalized to mean expression was equal to 100 in control genotype mice in each region for each protein. Values are means ± SE; n = 4/group in A–C and 8 control and 4 CD-Rhcg-KO in D.

DISCUSSION

The current studies examine Rhcg's role in the renal response to hypokalemia through the use of mice with collecting duct-specific Rhcg deletion. Hypokalemia increased Rhcg expression in both intercalated cells and principal cells in the OMCD. Thus it was surprising to us that the absence of collecting duct Rhcg expression did not impair total renal ammonia excretion in response to K⁺ deficiency. However, we found that during hypokalemia, collecting duct Rhcg deletion caused increased expression of PDG, the primary ammoniagenic enzyme in the kidney. Thus increased renal PDG expression in hypokalemic CD-Rhcg-KO mice seems to be a compensatory response that enables the normal increase in renal ammonia excretion during hypokalemia despite Rhcg deletion.

Our first major observation is that increased ammonia excretion in response to hypokalemia in the mouse coincides with increased Rhcg expression. Previous studies in the rat (11) and the current studies in the mouse demonstrate dramatic changes in Rhcg expression. These changes include increased total Rhcg protein expression in the OMCD in both the rat (11) and the mouse OMCD (current study) and increased both apical and basolateral Rhcg expression in OMCD cells in both the rat and the mouse (11 and current study). Thus these findings are consistent with Rhcg mediating an important role in the renal response to hypokalemia.

Several gene-deletion studies have demonstrated a key role for Rhcg in renal ammonia transport and urinary ammonia excretion. Both global and collecting duct-specific Rhcg deletion decrease basal ammonia excretion (3, 17). Metabolic acidosis is a potent stimulus for increased ammonia excretion, and global, collecting duct-specific, and intercalated cell-specific Rhcg deletions impair metabolic acidosis-stimulated ammonia excretion (3, 17, 18). In contrast, the increase in urinary ammonia excretion in response to hypokalemia is not altered by the absence of collecting duct Rhcg expression. This observation could indicate either that Rhcg is uninvolved in renal ammonia excretion in response to hypokalemia or that adaptive responses compensate for its absence.

Although there was residual Rhcg expression in the DCT and connecting tubule (CNT) in the gene-deletion strategy used in the current study (18), this is unlikely to significantly alter the findings. In particular, the majority of adaptations induced by hypokalemia occur in the outer medulla (9, 13), and not in the cortex, suggesting that hypokalemia-induced changes in ammonia secretion occur primarily in the more distal epithelial segments, and not in the CNT and DCT. Moreover, we did not identify adaptive changes in Rhcg expression in the CNT and DCT in response to hypokalemia, and did not identify differences in expression in these segments in response to CD-Rhcg-KO. Moreover, because the connecting segment as a site in which pendrin-mediated HCO₃⁻ secretion likely contributes to luminal alkalinization and the increased urine pH observed in response to hypokalemia, and increased urine pH in the segments likely decreases the net gradient for NH₃ secretion, it is very possible that ammonia secretion actually decreases in these segments in response to hypokalemia.

Our second major finding is that adaptations in proximal tubule ammoniagenesis appear to compensate for Rhcg deletion in hypokalemic states. Proximal tubule ammonia metabolism includes both ammoniagenesis and ammonia degradation. PEPCK and PDG are ammoniagenic enzymes present in the proximal tubule, and their expression and activity are increased by hypokalemia (5, 25, and current study). The observation that PDG expression in hypokalemic mice with CD-Rhcg-KO is greater than in hypokalemic mice with intact Rhcg expression suggests that enhanced ammoniagenesis via PDG, leading to increased interstitial ammonia accumulation, may compensate for Rhcg deletion in this condition.

NH₃ transport across plasma membranes involves multiple mechanisms. We have shown previously that apical ammonia transport in cultured collecting duct cells involves two distinct components, a transporter-mediated component, with functional characteristics identical to those defined for Rhcg, and a diffusive component of NH₃ movement (12). In perfused tubule studies, gene deletion studies have shown that Rhcg mediates some, but not all, of apical NH₃ transport (3). At present, the proportion of apical ammonia transport occurring through mechanisms other than Rhcg under basal conditions is incompletely understood. The only direct measurements of Rhcg's role in apical NH₃ transport examined perfused collecting duct segments from mice with metabolic acidosis, which has been shown to increase apical membrane Rhcg expression approximately fourfold in the rat (22). Importantly, only ∼60% of apical NH₃ permeability in collecting duct segments from mice with metabolic acidosis was dependent on apical Rhcg expression (3). If diffusive NH₃ movement mediates the residual, Rhcg-independent, NH₃ transport and if Rhcg-independent NH₃ transport was unchanged by metabolic acidosis, then ∼73% of baseline apical NH₃ permeability would occur through diffusive NH₃ transport and only ∼27% would involve Rhcg. Interestingly, these numbers are quite similar to the ∼30–35% reduction in baseline urinary ammonia excretion observed in mice with either total or collecting duct-specific Rhcg deletion (3, 17). Whether the diffusive component of apical plasma membrane NH₃ transport changes during physiological conditions has not been determined experimentally. However, hypokalemia increases the apical plasma membrane surface area of both intercalated and principal cells (9, 13), which is likely to increase diffusive NH₃ transport. Thus increased apical plasma membrane diffusive NH₃ transport is likely to occur in response to hypokalemia and may contribute in part to the observed increases in renal ammonia excretion.

Rhbg is another member of the ammonia transporter family and is expressed in the same renal epithelial cells as Rhcg, with the exception that its location is limited to the basolateral plasma membrane, whereas Rhcg is expressed in the apical and basolateral plasma membrane and in subapical vesicles. Our recent studies performed in parallel with the current study show that hypokalemia increases Rhbg expression and that intercalated cell-specific Rhbg expression is necessary for the normal increase in renal ammonia excretion by hypokalemia (1). This finding that Rhbg is important in the renal response to hypokalemia raises the possibility that increased Rhbg-mediated ammonia transport could compensate for the lack of Rhcg expression in the current study. However, neither total Rhbg expression nor Rhbg's cellular localization, as identifiable by immunohistochemistry, differed between hypokalemic mice with intact or with collecting duct-specific Rhcg deletion. We cannot exclude, however, alterations in either membrane targeting or phosphorylation, mechanisms shown in heterologous expression studies to be possible regulatory mechanisms of Rhbg (24).

The observations in the mice fed a diet with only 25% of normal K⁺ content provide interesting insights into the regulation of renal K⁺ and ammonia excretion. First, urinary K⁺ excretion decreased and urinary ammonia excretion increased significantly on the first day of the altered diet and remained altered during the 3 days of the experiment, even though serum K⁺ remained normal. Thus hypokalemia was not necessary for the changes in urinary ammonia and K⁺ excretion. Other studies have demonstrated that a K⁺-free diet for as little as 2 days alters urinary K⁺ and ammonia excretion despite the absence of hypokalemia (4); the current study shows that adaptive changes can occur even more quickly and do not require complete removal of K⁺ from the diet. One possible mechanism mediating changes in renal K⁺ excretion despite the absence of hypokalemia is through an intestinal kaliuretic hormone. Supporting this hypothesis is the observation that acute gastric K⁺ administration, before changes in serum K⁺, increases urinary K⁺ excretion and that acute administration of similar amounts of K⁺ directly into either peripheral veins or the portal vein does not elicit these acute changes (16). Importantly, changes in urinary ammonia and K⁺ excretion in response to altered dietary K⁺ intake do not require development of significant hypokalemia.

The current study, in combination with our parallel observations examining Rhbg's role in the response to hypokalemia, and previous work examining the response to metabolic acidosis, provides important insights into renal ammonia metabolism. In these parallel studies, deleting Rhbg in intercalated cells impaired hypokalemia-induced ammonia excretion (1), while deleting Rhcg from the entire collecting duct (current study) had a more modest effect. This contrasts with findings in metabolic acidosis, where the opposite was found: deleting Rhcg from either the entire genome or only from the collecting duct had a substantial effect on metabolic acidosis-induced ammonia excretion (3, 17), whereas the effect of intercalated cell-specific Rhbg deletion was more modest (2). These findings suggest that the mechanisms differ underlying the increased ammonia excretion in response to an acid load vs. potassium restriction.

In summary, the current studies provide important new information regarding the molecular mechanisms of renal ammonia metabolism during K⁺ restriction. Hypokalemia induced by dietary K⁺ deficiency increases Rhcg expression in the mouse kidney, particularly in the OMCD. Collecting duct-specific Rhcg deletion does not impair total renal ammonia excretion during hypokalemia, but this is likely due to enhanced renal ammoniagenesis via increased PDG expression in CD-Rhcg-KO mice, which would favor renal ammonia excretion via Rhcg-independent mechanisms. Finally, enhanced ammoniagenesis via increased PDG expression in the hypokalemic CD-Rhcg-KO mice is one mechanism that contributes to Rhcg-independent ammonia transport under these conditions and enables normal renal ammonia excretion despite the absence of collecting duct Rhcg.

GRANTS

These studies were supported by National Institutes of Health Grant DK-045788, the Merit Review Grant program of the Department of Veterans Affairs (1I01BX000818), and the National Research Foundation of Korea (2011-0016068). J. M. Bishop was supported by training grant DGE1011553 to Richard Snyder from the National Science Foundation.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: H.-W.L., J.M.B., and M.E.H. performed experiments; H.-W.L., J.W.V., J.M.B., M.E.H., K.-H.H., and I.D.W. analyzed data; H.-W.L. and I.D.W. drafted manuscript; H.-W.L., J.W.V., M.E.H., and I.D.W. approved final version of manuscript; J.W.V. and I.D.W. provided conception and design of research; J.W.V., K.-H.H., and I.D.W. interpreted results of experiments; J.W.V., J.M.B., M.E.H., K.-H.H., and I.D.W. edited and revised manuscript; I.D.W. prepared figures.