Abstract
Chronic metabolic acidosis results in renal Ca2+ and Mg2+ wasting, whereas chronic metabolic alkalosis is known to exert the reverse effects. It was hypothesized that these adaptations are mediated at least in part by the renal Ca2+ and Mg2+ transport proteins. The aim of this study, therefore, was to determine the effect of systemic acid-base status on renal expression of the epithelial Ca2+ channel TRPV5, the Ca2+-binding protein calbindin-D28K, and the epithelial Mg2+ channel TRPM6 in relation to Ca2+ and Mg2+ excretion. Chronic metabolic acidosis that was induced by NH4Cl loading or administration of the carbonic anhydrase inhibitor acetazolamide for 6 d enhanced calciuresis accompanied by decreased renal TRPV5 and calbindin-D28K mRNA and protein abundance in wild-type mice. In contrast, metabolic acidosis did not affect Ca2+ excretion in TRPV5 knockout (TRPV5−/−) mice, in which active Ca2+ reabsorption is effectively abolished. This demonstrates that downregulation of renal Ca2+ transport proteins is responsible for the hypercalciuria. Conversely, chronic metabolic alkalosis that was induced by NaHCO3 administration for 6 d increased the expression of Ca2+ transport proteins accompanied by diminished urine Ca2+ excretion in wild-type mice. However, this Ca2+-sparing action persisted in TRPV5−/− mice, suggesting that additional mechanisms apart from upregulation of active Ca2+ transport contribute to the hypocalciuria. Furthermore, chronic metabolic acidosis decreased renal TRPM6 expression, increased Mg2+ excretion, and decreased serum Mg2+ concentration, whereas chronic metabolic alkalosis resulted in the exact opposite effects. In conclusion, these data suggest that regulation of Ca2+ and Mg2+ transport proteins contributes importantly to the effects of acid-base status on renal divalent handling.
Acid-base homeostasis is known to affect renal handling of the divalents Ca2+ and Mg2+ (1–3). Chronic metabolic acidosis, which can occur as a result of clinical disorders such as renal failure, distal renal tubular acidosis, or chronic diarrhea, is associated with increased renal Ca2+ and Mg2+ excretion. Long-standing metabolic acidosis can lead to Ca2+ loss from bone and ultimately results in metabolic bone disease, including osteomalacia and osteoporosis (4). Conversely, chronic metabolic alkalosis in, for example, the milk-alkali syndrome, volume contraction, or treatment of nephrolithiasis by bicarbonate supplementation is known to decrease urine Ca2+ and Mg2+ excretion (1,3). However, the molecular mechanisms that explain the altered renal divalent excretion during these disturbances of acid-base balance remain unknown.
The major part of Ca2+ and Mg2+ reabsorption takes place in the proximal tubule and thick ascending limb of the loop of Henle (TAL) through a passive paracellular pathway (1,3). Fine-tuning of divalent excretion by the kidney occurs in the distal convoluted tubule (DCT) and the connecting tubule (CNT). In the latter segments, active transcellular Ca2+ and Mg2+ reabsorption determines the final amount excreted into the urine. Active Ca2+ reabsorption consists of Ca2+ entry through the apically localized epithelial Ca2+ channel TRPV5, cytosolic transport bound to the calcium-binding and buffering protein calbindin-D28K, and basolateral extrusion by the Na+/Ca2+ exchanger (NCX1) and a plasma membrane Ca2+ ATPase (PMCA1b) (1,5,6). Studying the regulation of active Mg2+ reabsorption in DCT has been seriously hampered by the lack of identification of the proteins involved (3). TRPM6 was recently identified as a Mg2+ permeable channel predominantly expressed along the apical membrane of DCT (7). Mutations in TRPM6 cause autosomal recessive hypomagnesemia, characterized by inappropriately high fractional Mg2+ excretion rates and disturbed intestinal Mg2+ absorption (8,9). This clearly suggests that TRPM6 constitutes the apical entry step in active Mg2+ (re)absorption and thus provides an important new tool to study this process at the molecular level (7).
Earlier studies, including micropuncture experiments, suggested that systemic acid-base disturbances specifically influence Ca2+ and Mg2+ reabsorption in DCT/CNT (2,10). Hypothetically, regulation of Ca2+ and Mg2+ transport proteins in these nephron segments could be involved in the altered renal divalent excretion secondary to changes in acid-base status, as was shown for other renal transporters and channels (11–14). We previously demonstrated that tacrolimus (FK506)-induced Ca2+ and Mg2+ wasting as well as thiazide-induced hypomagnesemia are associated with decreased renal expression of Ca2+ and/or Mg2+ transporters (15,16). Recently, TRPV5 knockout (TRPV5−/−) mice were generated in our laboratory, which display a robust renal Ca2+ leak localized to DCT/CNT, illustrating that active Ca2+ reabsorption is effectively abolished (17). These mice constitute a unique mouse model to determine the role of TRPV5 and active Ca2+ reabsorption in acid-base induced alterations of calciuresis.
The aim of this study, therefore, was to determine the effect of chronic metabolic acidosis and alkalosis on the expression of Ca2+ and Mg2+ transporters in the kidney and to evaluate their contribution to the altered Ca2+ and Mg2+ excretion. We induced metabolic alkalosis by oral NaHCO3 loading and metabolic acidosis by NH4Cl loading, as well as applied acetazolamide administration in wild-type and TRPV5−/− mice. Acetazolamide specifically inhibits proximal tubular HCO3 reabsorption, resulting in a self-limiting metabolic acidosis with, in contrast to NH4Cl loading, an alkaline urine pH (18–20). This enabled evaluation of the role of luminal pH. Furthermore, whereas acidosis generally increases urine Mg2+ excretion, acetazolamide is known for its unexplained Mg2+-sparing action (21–23).
Materials and Methods
Metabolic Acidosis and Alkalosis in Wild-Type and TRPV5−/− Mice
Metabolic Acidosis.
TRPV5−/− mice were recently generated by targeted ablation of the TRPV5 gene and genotyped as described previously (17). Ten-week-old wild-type (TRPV5+/+) mice and TRPV5−/− littermates were kept in a light- and temperature-controlled room with ad libitum access to deionized drinking water. Mice were ration-fed standard pelleted chow (0.25% [wt/wt] NaCl, 1.1% [wt/wt] Ca, 0.2% [wt/wt] Mg) during the metabolic balance studies. For evaluation of the effects of metabolic acidosis, mice were randomly assigned to a group that received either 0.28 M (TRPV5+/+) or 0.14 M (TRPV5−/−) NH4Cl via the drinking water during 6 d or a control group that received normal deionized drinking water. NH4Cl loading is a generally accepted and validated method to induce metabolic acidosis in rodents (11–13,24). Because the oral fluid intake of TRPV5−/− mice is approximately two-fold higher than that in wild-type mice, the lower NH4Cl concentration in their drinking water ensured a similar oral acid load compared with wild-type mice to prevent acid overloading in these TRPV5−/− mice. Alternatively, subcutaneous administration of the carbonic anhydrase inhibitor acetazolamide (20 mg/kg per d) during 6 d by osmotic minipumps was applied. This enabled evaluation of the role of urine pH in metabolic acidosis-induced changes in Ca2+ excretion as well as the role of TRPM6 in the unexplained Mg2+-sparing action of acetazolamide.
Metabolic Alkalosis.
Metabolic alkalosis was induced by oral administration of 0.2 and 0.1 M NaHCO3 to TRPV5+/+ and TRPV5−/− mice, respectively. NaHCO3 loading was previously shown to induce metabolic alkalosis in rodents (12,13). Mice that received 0.2 or 0.1 M NaCl constituted the control group, thereby correcting for possible effects of the increased Na+ load. The latter is particularly important because passive Ca2+ reabsorption is functionally coupled to Na+ reabsorption. Mice were treated for 6 d, after which they were housed in metabolic cages to enable collection of 24-h urine samples under mineral oil, preventing evaporation. At the end of the experiment, the mice were killed, blood samples were taken, and kidneys were sampled. The animal ethics board of the Radboud University Nijmegen approved all animal studies.
Analytical Procedures
Serum and urine Ca2+ and Mg2+ concentrations were determined using colorimetric assays as described previously (16,25). Blood gas measurements were performed using a Hitachi auto-analyzer (Hitachi, Laval, Quebec, Canada). Na+, K+, and Li+ concentrations were measured flame-spectrophotometrically (Eppendorf FCM 6343, Hamburg, Germany). Urine pH was determined using an electronic ion analyzer (Hanna Instruments, Szeged, Hungary), and osmolarity was measured with an Osmette A automatic osmometer (Precision Instruments, Sudbury, MA).
Real-Time Quantitative PCR
Total RNA was extracted from kidney using TriZol Total RNA Isolation Reagent (Life Technologies BRL, Breda, The Netherlands). The obtained RNA was subjected to DNAse treatment and reverse transcribed using Molony-Murine Leukemia Virus-Reverse Transcriptase (Life Technologies BRL) as described previously (26,27). Subsequently, the acquired cDNA was used to determine TRPV5, calbindin-D28K, and TRPM6 mRNA levels in kidney by real-time quantitative PCR on an ABI Prism 7700 Sequence Detection System (PE Biosystems, Rotkreuz, Switzerland) as described previously (15,26). In addition, mRNA expression of the housekeeping gene hypoxanthine-guanine phosphoribosyl transferase was determined as an endogenous control, which enabled calculation of specific mRNA expression levels as a ratio of hypoxanthine-guanine phosphoribosyl transferase.
Immunohistochemistry
Staining of kidney sections for TRPV5, calbindin-D28K, and TRPM6 was performed on cryosections of periodate-lysine-paraformaldehyde–fixed kidney samples as described previously (7,28). For semiquantitative determination of protein abundance, images were made using a Zeiss fluorescence microscope equipped with a digital camera (Nikon DXM1200), which were analyzed with the Image Pro Plus 4.1 image analysis software (Media Cybernetics, Silver Spring, MD). The entire cortex in two separate kidney sections of each animal was included in the analysis, resulting in quantification of protein levels as the mean of integrated optical density.
Immunoblotting
Calbindin-D28K protein levels were semiquantified by immunoblotting as described previously (27). In short, kidney cortex sections were homogenized and samples were normalized according to protein concentration. Subsequently, protein samples were separated on 16.5% (wt/vol) SDS-PAGE gels and blotted to polyvinylidene difluoride–nitrocellulose membranes (Immunobilon-P; Millipore Corp., Bedford, MA), and protein was detected using a rabbit calbindin-D28K antibody.
Statistical Analyses
Data are expressed as means ± SEM. Statistical comparisons were analyzed by one-way ANOVA and Fisher multiple comparison. P < 0.05 was considered statistically significant. All analyses were performed using the StatView Statistical Package software (Power PC version 4.51, Berkely, CA) on an Apple iMac computer.
Results
Metabolic Acidosis and Alkalosis in Wild-Type and TRPV5−/− Mice
Oral NH4Cl loading induced a similar metabolic acidosis in wild-type and TRPV5−/− mice, as demonstrated by the significantly reduced blood pH and HCO3− concentration (Table 1). Accordingly, NH4Cl reduced urine pH in wild-type and TRPV5−/− mice compared with their respective controls (Table 2). Likewise, chronic acetazolamide treatment significantly increased urine pH and decreased blood pH and HCO3− concentration in wild-type but not in TRPV5−/− mice. Of note, acetazolamide-induced acidosis and urinary alkalinization is often self-limiting. Indeed, urine pH was more alkaline in all animals after 3 d compared with day 6 of treatment (data not shown). Furthermore, blood gas determination substantiated the effectiveness of the oral NaHCO3 loading protocol in wild-type and TRPV5−/− mice. Serum pH was similar in both genotypes during NaHCO3 loading, whereas serum HCO3− levels in TRPV5−/− mice were significantly higher. This reflects higher pCO2 in the latter situation, most probably as a result of differences in time or depth of anesthesia. The absence of this difference in pCO2 between wild-type and TRPV5−/− mice in the other treatment groups suggests that there is no intrinsic increased susceptibility to retain CO2 in TRPV5−/− mice. It is interesting that urine pH was consistently 0.5 to 1 pH unit lower in TRPV5−/− mice as compared with the corresponding wild-type mice, but, importantly, TRPV5−/− mice did not display metabolic acidosis at baseline. Diuresis and Na+ excretion were not affected by NH4Cl or NaHCO3 loading, whereas acetazolamide significantly increased urine volume and natriuresis in wild-type and TRPV5−/− mice (Table 2).
Acid-base status during different treatment protocols in TRPV5+/+ and TRPV5−/− micea
Urine composition and fluid intake during chronic metabolic acidosis and alkalosis in TRPV5+/+ and TRPV5−/− micea
Ca2+ Homeostasis during Chronic Metabolic Acidosis and Alkalosis
Genetic ablation of TRPV5 resulted in a strikingly increased calciuresis compared with wild-type littermates (Figure 1A). NH4Cl loading significantly enhanced urine Ca2+ excretion in wild-type mice, whereas Ca2+ excretion was not affected in TRPV5−/− mice. Likewise, acetazolamide treatment significantly enhanced calciuresis in wild-type mice, whereas this effect was not present in TRPV5−/− mice. Serum Ca2+ levels remained unaltered during NH4Cl loading and acetazolamide treatment (Figure 1C). In contrast to metabolic acidosis, NaHCO3 administration significantly reduced urine Ca2+ excretion in wild-type as well as in TRPV5−/− mice (Figure 1B). Serum Ca2+ levels and urine volume did not differ between the alkalosis and control groups (Figure 1D, Table 1). Because Li+ and Na+ are transported in parallel by the proximal tubule, endogenous Li+ clearance was used as an inverse measure of proximal tubular Na+ reabsorption, to which in turn passive Ca2+ reabsorption is functionally coupled (16). Li+ clearance was significantly increased by acetazolamide treatment and NaCl loading in wild-type and TRPV5−/− mice, suggesting decreased proximal tubular Na+ reabsorption (Table 1). Conversely, NaHCO3 loading decreased Li+ clearance compared with NaCl-treated controls.
Urinary Ca2+ excretion and serum Ca2+ concentration during metabolic acidosis and alkalosis in wild-type (TRPV5+/+) and TRPV5 knockout (TRPV5−/−) mice. The effects of chronic metabolic acidosis and acetazolamide (A) as well as during metabolic alkalosis (B) on renal Ca2+ excretion and serum Ca2+ concentration (C and D) were determined in metabolic cage experiments (n = 9 animals; n = 3 animals per cage). Controls, animals that received deionized drinking water only; NH4Cl, animals that received 0.28 M (TRPV5+/+) or 0.14 M (TRPV5−/−) NH4Cl via the drinking water; Acetazolamide, animals that received acetazolamide (20 mg/kg per d) subcutaneously by osmotic minipump; NaCl, animals that received 0.2 (TRPV5+/+) or 0.1 M (TRPV5−/−) NaCl via the drinking water; NaHCO3, animals that received 0.2 (TRPV5+/+) or 0.1 M (TRPV5−/−) NaHCO3 via the drinking water. Data are presented as means ± SEM. *P < 0.05 versus respective TRPV5+/+ or TRPV5−/− control group (controls or NaCl-treated animals).
Mg2+ Homeostasis during Chronic Metabolic Acidosis and Alkalosis
In addition, we evaluated the effect of the different treatment protocols on renal Mg2+ homeostasis in wild-type mice. Oral NH4Cl loading significantly enhanced Mg2+ excretion (Figure 2A), which was accompanied by decreased serum Mg2+ levels (Figure 2C). In contrast, acetazolamide displayed a Mg2+-sparing effect accompanied by a significantly increased serum Mg2+ concentration. Likewise, metabolic alkalosis that was induced by NaHCO3 treatment significantly reduced urine Mg2+ excretion (Figure 2B) and increased the serum Mg2+ level (Figure 2D).
Urinary Mg2+ excretion and serum Mg2+ concentration during metabolic acidosis and alkalosis in wild-type mice. The effects of chronic metabolic acidosis and acetazolamide (A) as well as during metabolic alkalosis (B) on renal Mg2+ excretion and serum Mg2+ concentration (C and D) were determined in metabolic cage experiments (n = 9 animals; n = 3 animals per cage). Controls, animals that received deionized drinking water only; NH4Cl, animals that received 0.28 M NH4Cl via the drinking water; Acetazolamide, animals that received acetazolamide (20 mg/kg per d) subcutaneously by osmotic minipump; NaCl, animals that received 0.2 M NaCl via the drinking water; NaHCO3, animals that received 0.2 M NaHCO3 via the drinking water. Data are presented as means ± SEM. *P < 0.05 versus respective control group (controls or NaCl-treated animals).
Renal mRNA and Protein Expression of Ca2+ Transporters
For studying the effect of systemic acid-base status on renal Ca2+ transporter expression, TRPV5 and calbindin-D28K mRNA levels were determined by real-time quantitative PCR analysis, and protein abundance was analyzed by immunohistochemistry and immunoblotting. NH4Cl loading significantly reduced both TRPV5 and calbindin-D28K mRNA levels in kidney cortex of wild-type mice (Figure 3, A and C). Calbindin-D28K mRNA levels were significantly decreased in TRPV5−/− compared with TRPV5+/+ mice. In addition, NH4Cl treatment further reduced calbindin-D28K mRNA levels compared with control TRPV5−/− mice. Figure 4A shows representative immunohistochemical images of kidney cortex probed with anti-TRPV5 and anti–calbindin-D28K antibodies. In addition, calbindin-D28K protein abundance was determined by immunoblotting (Figure 4B). In line with the mRNA levels, semiquantification of the immunohistochemical analysis showed that NH4Cl decreases TRPV5 and calbindin-D28K protein abundance (Figure 5, A and C). These results were confirmed by immunoblotting (100 ± 4 versus 63 ± 9% in wild-type and 100 ± 5 versus 75 ± 8% in TRPV5−/− mice, respectively). Acetazolamide treatment decreased TRPV5 mRNA and protein expression in wild-type mice (Figures 3A and 5A). Furthermore, calbindin-D28K protein abundance was reduced in acetazolamide-treated mice (Figure 5C), as confirmed by immunoblotting (100 ± 1 versus 63 ± 6 and 40 ± 8% in wild-type and TRPV5−/− mice, respectively). In contrast to metabolic acidosis, chronic metabolic alkalosis that was induced by NaHCO3 loading increased TRPV5 mRNA and protein expression in wild-type mice (Figures 3B and 5B). Likewise, calbindin-D28K expression was increased as determined by real-time PCR and immunohistochemistry (Figures 3D and 5D), as well as by immunoblotting (100 ± 41 versus 293 ± 40%). Conversely, calbindin-D28K mRNA and protein levels were not significantly increased in NaHCO3-treated TRPV5−/− mice.
Effect of chronic metabolic acidosis and alkalosis on renal mRNA expression of Ca2+ transport proteins in TRPV5+/+ and TRPV5−/− mice. Renal mRNA expression levels of the epithelial Ca2+ channel TRPV5 and the cytosolic Ca2+-binding protein calbindin-D28K (CaBP28K) were determined during chronic metabolic acidosis and acetazolamide treatment (A and C, respectively) and chronic metabolic alkalosis (B and D, respectively) by real-time quantitative PCR analysis as the ratio of hypoxanthine-guanine phosphoribosyl transferase (HPRT) and depicted as percentage of respective controls. Controls, animals that received deionized drinking water only; NH4Cl, animals that received 0.28 (TRPV5+/+) or 0.14 M (TRPV5−/−) NH4Cl via the drinking water; Acetazolamide, animals that received acetazolamide (20 mg/kg per d) subcutaneously by osmotic minipump; NaCl, animals that received 0.2 (TRPV5+/+) or 0.1 M (TRPV5−/−) NaCl via the drinking water; NaHCO3, animals that received 0.2 (TRPV5+/+) or 0.1 M (TRPV5−/−) NaHCO3 via the drinking water; n = 9 animals per group. Data are presented as means ± SEM. *P < 0.05 versus respective TRPV5+/+ or TRPV5−/− control group (controls or NaCl-treated animals); #P < 0.05 versus TRPV5+/+ controls.
Immunohistochemical staining and semiquantitative immunoblotting of renal Ca2+ transport proteins in TRPV5+/+ mice. (A) Representative immunohistochemical images of TRPV5 and CaBP28K staining in kidney cortex of wild-type mice. (B) Representative immunoblots for calbindin-D28K. Controls, animals that received deionized drinking water only; NH4Cl, animals that received 0.28 M NH4Cl via the drinking water; ACTZ, animals that received acetazolamide (20 mg/kg per d) subcutaneously by osmotic minipump; NaCl, animals that received 0.2 M NaCl via the drinking water; NaHCO3, animals that received 0.2 M NaHCO3 via the drinking water.
Effect of chronic metabolic acidosis and alkalosis on renal protein abundance of Ca2+ transport proteins in TRPV5+/+ and TRPV5−/− mice. Renal protein expression levels of the epithelial Ca2+ channel TRPV5 and the cytosolic CaBP28K were determined during chronic metabolic acidosis and acetazolamide treatment (A and C, respectively) as well as chronic metabolic alkalosis (B and D, respectively) by computerized analysis of immunohistochemical images. Data were calculated as integrated optical density (IOD; arbitrary units) and depicted as percentage of respective controls. Controls, animals that received deionized drinking water only; NH4Cl, animals that received 0.28 (TRPV5+/+) or 0.14 M (TRPV5−/−) NH4Cl via the drinking water; Acetazolamide, animals that received acetazolamide (20 mg/kg per d) subcutaneously by osmotic minipump; NaCl, animals that received 0.2 (TRPV5+/+) or 0.1 M (TRPV5−/−) NaCl via the drinking water; NaHCO3, animals that received 0.2 (TRPV5+/+) or 0.1 M (TRPV5−/−) NaHCO3 via the drinking water; n = 9 animals per treatment group. Data are presented as means ± SEM. *P < 0.05 versus respective TRPV5+/+ or TRPV5−/− control group (controls or NaCl-treated animals); #P < 0.05 versus TRPV5+/+ controls.
Effect of Acid-Base Status on Renal TRPM6 Expression
Renal TRPM6 mRNA and protein expression levels were determined by real-time quantitative PCR analysis (Figure 6) and semiquantitative immunohistochemistry (Figure 7A). Both NH4Cl loading and acetazolamide treatment significantly reduced renal TRPM6 mRNA as well as protein abundance in wild-type mice (Figures 6A and 7B), whereas NaHCO3-treated mice displayed increased TRPM6 expression (Figures 6B and 7C).
Effect of chronic metabolic acidosis and alkalosis on renal mRNA expression of the epithelial Mg2+ channel TRPM6 in wild-type mice. Renal mRNA expression levels of TRPM6 was determined during chronic metabolic acidosis and acetazolamide treatment (A) as well as chronic metabolic alkalosis (B) by real-time quantitative PCR analysis as the ratio of HPRT and depicted as percentage of respective controls. Controls, animals that received deionized drinking water only; NH4Cl, animals that received 0.28 M NH4Cl via the drinking water; Acetazolamide, animals that received acetazolamide (20 mg/kg per d) subcutaneously by osmotic minipump; NaCl, animals that received 0.2 M NaCl via the drinking water; NaHCO3, animals that received 0.2 M NaHCO3 via the drinking water; n = 9 animals per treatment group. Data are presented as means ± SEM. *P < 0.05 versus respective control group (controls or NaCl-treated animals).
Effects of chronic metabolic acidosis and alkalosis on renal protein abundance of the epithelial Mg2+ channel TRPM6 in wild-type mice. Representative immunohistochemical imaging of TRPM6 staining in kidney cortex (A), which enabled semiquantitative determination of renal TRPM6 protein abundance during chronic metabolic acidosis and acetazolamide treatment (B) as well as chronic metabolic alkalosis (C) by computerized analysis. Data were calculated as IOD (arbitrary units) and depicted as percentage of respective controls. Controls, animals that received deionized drinking water only; NH4Cl, animals that received 0.28 M NH4Cl via the drinking water; ACTZ/Acetazolamide, animals that received acetazolamide (20 mg/kg per d) subcutaneously by osmotic minipump; NaCl, animals that received 0.2 M NaCl via the drinking water; NaHCO3, animals that received 0.2 M NaHCO3 via the drinking water; n = 9 animals per treatment group. Data are presented as means ± SEM. *P < 0.05 versus respective control group (controls or NaCl-treated animals).
Discussion
This study demonstrated that systemic acid-base status regulates the expression of proteins that are involved in active Ca2+ and Mg2+ reabsorption. Our data showed that downregulation of renal Ca2+ transport proteins is responsible for the hypercalciuria during chronic metabolic acidosis. In contrast, the Ca2+-sparing effect of chronic metabolic alkalosis was associated with enhanced Ca2+ transporter abundance. However, the Ca2+-sparing action persisted in TRPV5−/− mice, suggesting that additional mechanisms apart from upregulation of active Ca2+ transport contribute to the hypocalciuria. Furthermore, metabolic acidosis decreased renal TRPM6 abundance as well as Mg2+ reabsorption, whereas metabolic alkalosis had the opposite effect. These data indicate that regulation of TRPM6 explains the effects of acid-base status on renal Mg2+ handling.
Chronic metabolic acidosis that was induced by NH4Cl loading enhanced Ca2+ excretion and decreased the expression of the epithelial Ca2+ channel TRPV5 and the cytosolic Ca2+-binding and buffering protein calbindin-D28K in wild-type mice. Both proteins play a central role in active Ca2+ reabsorption in DCT/CNT (1). We showed that 0.14 M NH4Cl loading induced a similar metabolic acidosis in polydipsic TRPV5−/− mice compared with wild-type mice. Importantly, Ca2+ excretion was not altered during chronic metabolic acidosis in TRPV5−/− mice, in which active Ca2+ reabsorption is effectively abolished (17). These results indicated that downregulation of Ca2+ transport proteins that are present in DCT/CNT underlies the increased Ca2+ excretion during NH4Cl loading. Alternatively, increased Ca2+ mobilization from bone has been shown in chronic metabolic acidosis and was suggested to explain the Ca2+ wasting (4,29). Our study provides evidence for a primary renal Ca2+ leak. In line with our data, previous micropuncture experiments suggested that tubular Ca2+ reabsorption in DCT/CNT is specifically diminished during chronic metabolic acidosis in dogs (2). In contrast, Rizzo et al. (30) previously reported that NH4Cl-induced acidosis in rats was accompanied by a moderate although significant increase of calbindin-D28K. The reason for the discrepancy with our study, which shows a consistent decrease of TRPV5 and calbindin-D28K mRNA as well as protein expression, is not known.
Mice that receive NH4Cl develop chronic metabolic acidosis that is characterized by a slight decrease in blood pH, a reduced serum HCO3− concentration, and, importantly, a low urine pH as substantiated in this study (24). Vennekens et al. (31) recently demonstrated that extracellular protons inhibit TRPV5 in vitro by titrating glutamate 522 in the extracellular loop between the fifth putative transmembrane domain and the pore region as shown by Yeh et al. (32). Therefore, it was suggested that acidification of the DCT/CNT luminal fluid during chronic metabolic acidosis explains the decreased Ca2+ reabsorption in vivo (31,32). These data raised the question of whether the regulation of Ca2+ transport proteins is secondary to the acidosis per se or due to the low urine pH. This urine acidification has been attributed mainly to parallel enhancement of apical Na+/H+ exchanger (NHE3) and basolateral Na+-HCO3− co-transporter activity in the proximal tubule (33–36). In contrast, acetazolamide treatment is known to induce metabolic acidosis by diminishing this proximal tubular HCO3− reabsorptive capacity (18–20). This would result in an increased luminal pH at more distal nephron segments, including TRPV5 and calbindin-D28K–expressing DCT/CNT. In our study, acetazolamide-treated mice indeed displayed urine alkalinization. Importantly, acetazolamide downregulated the expression of Ca2+ transport proteins. Thus, luminal pH in DCT/CNT does not seem to be crucial in the long-term in vivo hypercalciuric effect of chronic metabolic acidosis. Therefore, our data provide a molecular explanation for the increased Ca2+ excretion in clinically relevant situations, including chronic renal failure, chronic diarrhea, and renal tubular acidosis. In particular, acetazolamide treatment is a common cause of proximal renal tubular acidosis and is often associated with Ca2+ nephrolithiasis (37,38). Taken together, we showed that systemic metabolic acidosis, as opposed to associated changes in urine pH, downregulates Ca2+ transport proteins and, thereby, induces hypercalciuria.
Conversely, chronic metabolic alkalosis is known to decrease Ca2+ excretion (2). In this study, chronic NaHCO3 administration induced metabolic alkalosis in wild-type as well as TRPV5−/− mice. Chronic metabolic alkalosis had a Ca2+-sparing effect and increased renal expression of the Ca2+ transport proteins in wild-type mice. This suggests that the increased expression of these transporters is responsible for the hypocalciuric effect. However, TRPV5 ablation resulting in the functional lack of active Ca2+ transport did not preclude this effect, suggesting that upregulation of Ca2+ transport proteins in DCT/CNT is not crucial for the induction of hypocalciuria. It is interesting that alkalosis did not enhance calbindin-D28K abundance in TRPV5−/− mice. This is in line with previous studies from our laboratory, which demonstrated that blockade of the TRPV5-mediated Ca2+ influx in rabbit CNT/CCD cells downregulates calbindin-D28K expression (39). This indicated that regulation of the latter protein is highly dependent on the presence of TRPV5. The bulk of filtered Ca2+ is reabsorbed by a passive paracellular mechanism that is localized primarily in the proximal tubule and to a lesser extent in TAL (1). In these nephron segments, Ca2+ reabsorption is secondary to Na+ reabsorption and the resulting water reabsorption, which creates a favorable electrochemical gradient driving passive Ca2+ transport. Determination of Li+ clearance indeed suggested that NaHCO3-treated mice show increased proximal tubular Na+ reabsorption and, therefore, possibly display enhanced passive Ca2+ reabsorption compared with NaCl-treated controls. Thus, the hypocalciuria could alternatively be explained by increased passive Ca2+ reabsorption. Taken together, the present data offer insight into the previously unexplained mechanism by which administration of HCO3− prevents nephrolithiasis in patients with recurrent kidney stones (40,41). Upregulation of Ca2+ transport proteins in DCT/CNT occurs but is not crucial for the Ca2+-sparing effect.
The epithelial Mg2+ channel TRPM6 is the first identified protein involved in active Mg2+ reabsorption (7–9). TRPM6 was localized along the apical membrane of DCT, and mutations in the gene encoding TRPM6 were shown to cause autosomal recessive hypomagnesemia, characterized by inappropriately high Mg2+ excretion and disturbed intestinal Mg2+ absorption. In our study, NH4Cl-induced chronic metabolic acidosis decreased renal TRPM6 abundance accompanied by increased Mg2+ excretion and hypomagnesemia. Conversely, chronic metabolic alkalosis increased TRPM6 expression as well as renal Mg2+ reabsorption, resulting in hypermagnesemia. There is insufficient functional information available regarding the tubular segments that are involved in the altered Mg2+ reabsorption, but Wong et al. (10,42) previously demonstrated altered Mg2+ reabsorption in the distal tubule during metabolic acidosis and alkalosis in the dog. Furthermore, a high extracellular pH was shown to enhance Mg2+ uptake in isolated mouse DCT cells, and conversely a low pH diminished this uptake (43). Thus, these data suggest that alterations of acid-base status regulate TRPM6 expression, thereby affecting renal active Mg2+ reabsorption in DCT and leading to significant changes in serum Mg2+. We previously demonstrated that thiazide administration as well as treatment with the hypomagnesemic immunosuppressant tacrolimus (FK506) reduces renal TRPM6 abundance accompanied by increased urine Mg2+ loss (15,16). Therefore, TRPM6 downregulation seems to be a general mechanism explaining the renal Mg2+ leak and resulting hypomagnesemia in these important clinical situations.
It is interesting that we showed that acetazolamide displays a Mg2+-sparing effect in mice. This carbonic anhydrase inhibitor is also known to result in hypomagnesuria in human via an unknown mechanism (21–23). Importantly, TRPM6 expression was significantly diminished during chronic acetazolamide treatment. This suggests that chronic metabolic acidosis, irrespective of cause or associated urine pH, downregulates TRPM6 expression. At the same time, these data are against enhanced active Mg2+ reabsorption explaining the decreased Mg2+ excretion. In contrast to Ca2+, the bulk of filtered Mg2+ is reabsorbed in TAL, where the Na+-K+(NH4+)-2Cl− (NKCC2) cotransporter is responsible for maintaining the required electrochemical gradient (44). Acetazolamide treatment was associated with increased urine volume and Na+ excretion, which has been shown to result in extracellular volume contraction and enhanced NKCC2 activity (45,46). Furthermore, metabolic acidosis, which developed in acetazolamide-treated wild-type mice, was previously shown to increase NKCC2 mRNA and protein abundance (14,47). Thus, we postulate that these additive stimulatory effects enhance passive Mg2+ reabsorption in TAL. Alternatively, regulation of the tight junction protein paracellin-1, which is supposed to facilitate paracellular reabsorption of Mg2+ in TAL, might be involved (48,49). Together, these mechanisms could counteract the metabolic acidosis-induced TRPM6 downregulation and result in a net Mg2+-sparing effect.
The mechanism translating the acid-base status to regulation of gene expression remains largely unknown. Our results show that transcriptional regulation occurs irrespective of urine pH. Because intracellular pH will ultimately reflect pH of the urine, apical or intracellular acid sensing does not seem to be involved. Therefore, direct sensing of acid-base status by pH-sensitive proteins at the basolateral membrane is likely. In DCT/CNT, the extracellular Ca2+/Mg2+-sensing receptor is expressed at the basolateral membrane (50,51). It was shown recently that extracellular pH directly influences the sensitivity of this receptor to Ca2+ and Mg2+ (52,53). Therefore, altered Ca2+/Mg2+ sensing might influence transcellular divalent transport in these nephron segments. Furthermore, other basolateral proton-sensing receptors or channels might act as an acid sensor regulating Ca2+ and Mg2+ transport protein expression (54,55). Thus, whereas the presented data demonstrated the transcriptional and translational regulation of proteins that are involved in active Ca2+ and Mg2+ reabsorption, future studies will have to elucidate the exact mechanism by which acid-base status affects expression of these transporters.
Acknowledgments
This work was financially supported by the Dutch Kidney Foundation (C10.1881, C03.6017) and the Dutch Organization of Scientific Research (Zon-Mw 016.006.001).
We thank the Central Animal Facility, Radboud University Nijmegen, for technical support.
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
- © 2006 American Society of Nephrology