Mg²⁺ restriction downregulates NCC through NEDD4-2 and prevents its activation by hypokalemia

Abstract

Hypomagnesemia is associated with reduced kidney function and life-threatening complications and sustains hypokalemia. The distal convoluted tubule (DCT) determines final urinary Mg²⁺ excretion and, via activity of the Na⁺-Cl⁻ cotransporter (NCC), also plays a key role in K⁺ homeostasis by metering Na⁺ delivery to distal segments. Little is known about the mechanisms by which plasma Mg²⁺ concentration regulates NCC activity and how low-plasma Mg²⁺ concentration and K⁺ concentration interact to modulate NCC activity. To address this, we performed dietary manipulation studies in mice. Compared with normal diet, abundances of total NCC and phosphorylated NCC (pNCC) were lower after short-term (3 days) or long-term (14 days) dietary Mg²⁺ restriction. Altered NCC activation is unlikely to play a role, since we also observed lower total NCC abundance in mice lacking the two NCC-activating kinases, STE20/SPS-1-related proline/alanine-rich kinase and oxidative stress response kinase-1, after Mg²⁺ restriction. The E3 ubiquitin-protein ligase NEDD4-2 regulates NCC abundance during dietary NaCl loading or K⁺ restriction. Mg²⁺ restriction did not lower total NCC abundance in inducible nephron-specific neuronal precursor cell developmentally downregulated 4-2 (NEDD4-2) knockout mice. Total NCC and pNCC abundances were similar after short-term Mg²⁺ or combined Mg²⁺-K⁺ restriction but were dramatically lower compared with a low-K⁺ diet. Therefore, sustained NCC downregulation may serve a mechanism that enhances distal Na⁺ delivery during states of hypomagnesemia, maintaining hypokalemia. Similar results were obtained with long-term Mg²⁺-K⁺ restriction, but, surprisingly, NCC was not activated after long-term K⁺ restriction despite lower plasma K⁺ concentration, suggesting significant differences in distal tubule adaptation to acute or chronic K⁺ restriction.

INTRODUCTION

Mg²⁺ is the second most abundant intracellular divalent cation and is essential for many cellular processes (16). It is required for activities of all ATP-dependent enzymes, many reactions involving kinases, regulation of mitochondrial function and ion channels, and neuromuscular excitability (15, 16, 28). As a result, hypomagnesemia frequently leads to neuromuscular hyperexcitability (ranging from tremors to convulsions) and neuropsychiatric disturbances (28). Hypomagnesemia is frequently associated with hypokalemia and/or hypocalcemia, which can result in cardiac arrhythmia and sudden death (5, 20). Chronic hypomagnesemia has adverse effects in many organs, including the kidney (48), and is associated with higher production of inflammatory and proatherogenic cytokines in endothelial cells that might lead to impaired renal function (37). In the setting of existing kidney disease, reduced plasma Mg²⁺ concentration may increase the risk of end-stage renal disease (68). A recent study (49) conducted in a cohort of healthy African-American and Caucasian adults showed that low dietary Mg²⁺ intake was associated with a rapid decline in kidney function.

Plasma Mg²⁺ concentration is strictly maintained by the interplay of intestinal absorption, renal transport, and bone exchange (67). About 80% of total plasma Mg²⁺ is filtered by the kidney, of which 90–95% is reabsorbed by various segments of the nephron (13). Most filtered Mg²⁺ undergoes passive paracellular reabsorption along the proximal tubule and thick ascending limb of the loop of Henle (6). Downstream of these segments is the distal convoluted tubule (DCT), which expresses the Na⁺-Cl⁻ cotransporter (NCC), the target of thiazide diuretics. The DCT reabsorbs only 5–10% of the filtered Mg²⁺ by active transcellular transport via apical transient receptor potential M6/7 channels but plays a key role determining the final urinary excretion of Mg²⁺ (6, 13). This is demonstrated by the presentation of hypomagnesemia as a central feature of Gitelman syndrome caused by recessive mutations in slc12a3, which encodes NCC (59). Furthermore, pharmacological blockade of NCC by thiazides or genetic disruption of NCC in mice also causes hypomagnesemia (57).

The DCT also plays a central role in K⁺ homeostasis by acting as a sensor of plasma K⁺ concentration, which modulates NCC phosphorylation status (61, 62). When plasma K⁺ concentration decreases, K⁺ exits DCT cells through basolateral Kir4.1 channels; the resultant membrane depolarization increases Cl⁻ efflux through ClC-Kb channels (62, 75, 76). Lower intracellular Cl⁻ then disinhibits the Cl⁻-sensitive kinase with no lysine (K) kinase-4 (WNK4), leading to phosphorylation and activation of STE20/SPS-1-related proline/alanine-rich kinase (SPAK) and oxidative stress response kinase-1 (OSR1) (61, 62). SPAK/OSR1 directly phosphorylates the NH₂-terminus of NCC, which activates NCC and decreases delivery of Na⁺ to distal segments. Because Na⁺ delivery generates the electrogenic drive for K⁺ secretion by distal segments, this results in reduced K⁺ secretion and maintenance of plasma K⁺ (3, 39, 60, 79). This dual role of the DCT in Mg²⁺ and K⁺ homeostasis is likely to be clinically important. Hypomagnesemia is one of the most frequently encountered forms of electrolyte dysregulation in critically ill patients; the prevalence is ∼10% of hospitalized patients and can be as high as 65% in intensive care patients (11, 55). Thiazide diuretics, which directly block NCC, cause both K⁺ and Mg²⁺ deficiencies (74). Importantly, Mg²⁺ deficiency exacerbates the effect of hypokalemia; hypokalemia with concomitant Mg²⁺ deficiency is often refractory to K⁺ supplement. Coadministration of Mg²⁺ is essential to rectify hypokalemia (29).

Many studies have shown that dietary interventions in addition to dietary K⁺ restriction have a robust influence on NCC activity (62). A K⁺-rich diet rapidly dephosphorylates and thereby deactivates NCC in a Cl⁻-and SPAK/OSR1-independent manner (47, 58). NCC abundance and phosphorylation are increased by low NaCl intake and decreased by high NaCl intake (12). Serum and glucocorticoid-inducible kinase 1 (SGK1) has been implicated in mediating low NaCl-mediated activation of NCC through the WNK-SPAK/OSR1 signaling pathway (65). A recent study (69) showed that mice displaying hyperactivation of NCC, induced by expression of constitutively active SPAK along the DCT, are normomagnesemic, suggesting an uncoupling of Na⁺ and Mg²⁺ reabsorption, at least in states of elevated NCC activity. In contrast, very little is known regarding how changes in plasma Mg²⁺ concentration modify NCC activity and DCT function.

One study (66) reported the effect of modifying plasma Mg²⁺ concentration on magnesiotropic genes, including slc12a3 in mice by chronically (14 days) providing Mg²⁺-deficient and high-Mg²⁺ diets. Whereas NCC mRNA levels did not differ, immunohistochemistry revealed lower abundance of total NCC after Mg²⁺ restriction compared with high-Mg²⁺ diet (66). However, this study focused only on total NCC and did not evaluate NCC phosphorylation, a key determinant of total NCC activity and stability (41, 54, 66). The mechanisms leading to lower NCC abundance and whether altered activity of the WNK-SPAK/OSR1 pathway plays a role have also not been determined. Finally, it is unknown whether the effects of dietary Mg²⁺ restriction on NCC occur rapidly, as occurs with changes in dietary K⁺, or whether there is an interplay between Mg²⁺ and K⁺ restriction. This is clinically relevant, since it has been proposed that increased Na⁺ delivery to K⁺-secreting segments, which may occur with lower NCC activity after Mg²⁺ restriction, is an important component of hypokalemia in the setting of hypomagnesemia (29). Therefore, in the present study, we investigated 1) the acute and chronic effects of dietary Mg²⁺ manipulation on total NCC and phosphorylated NCC (pNCC), 2) the contribution of the WNK-SPAK/OSR1 pathway, 3) whether the E3 ubiquitin-protein ligase neuronal precursor cell developmentally downregulated 4-2 (NEDD4-2) previously shown to play a role in NCC degradation (18, 53) mediates the effects of Mg²⁺ restriction, and 4) whether downregulation of NCC by dietary Mg²⁺ restriction overrides the strong stimulatory effect of K⁺ restriction on NCC.

MATERIALS AND METHODS

Animals.

C57BL/6J mice aged 3–5 mo maintained in a temperature-controlled and 12:12-h light-dark cycle room were used in this study. To investigate the effect of dietary Mg²⁺, mice were fed a normal Mg²⁺ [0.15% (wt/wt); NaCl ∼0.49%, K⁺ ~0.8%], Mg²⁺-deficient [0.0015–0.003% (wt/wt); NaCl ∼0.49%, K⁺ ∼0.8%], high-Mg²⁺ [0.48% (wt/wt); NaCl ∼0.49%, K⁺ ∼0.8%], Mg²⁺-K⁺-deficient (NaCl ∼0.49%), or K⁺-deficient diet (Mg²⁺ ∼0.15%, NaCl ∼0.49%) for 3 days (short-term study) or 14 days (long-term study). Normal and high-Mg²⁺ diets were prepared by adding MgO [to 0.15% (wt/wt) Mg²⁺] to a Mg²⁺-deficient diet (Teklad Custom Diet, TD.170002, Envigo). In short-term and long-term experiments to investigate the effect of dietary Mg²⁺ and K⁺ deficiencies, Mg²⁺-K⁺-deficient diet (Teklad Custom Diet, TD.170943, Envigo) was used. K⁺-deficient diet was prepared by adding MgO [to 0.15% (wt/wt) Mg²⁺]. To prepare normal or Mg²⁺-deficient diets, potassium citrate monohydrate [to 0.58% (wt/wt)], potassium sulfate [to 0.22% (wt/wt)], chromium potassium sulfate dodecahydrate [to 0.0001% (wt/wt)], and potassium iodate (1 N, 0.18 µl/100 g) were added to the Mg²⁺-K⁺-deficient diet, with the normal diet also having MgO added. All knockout (KO) mice were bred in house and were on a C57bl/6 background. The Pax8-rtTA/LC1 system was used to generate inducible renal epithelia-specific KO mice (64). SPAK and OSR1 double-KO (DKO) mice and OSR1^fl/fl mice were bred with SPAK-KO mice until homozygous OSR1^fl/fl/SPAK KO mice were obtained, which were then bred with Pax8-rtTA/LC1 mice. The resulting offspring were interbred to finally obtain OSR1^fl/fl/SPAK KO/Pax8-rtTA/LC1 mice (21). Inducible nephron-specific NEDD4-2 KO mice were generated by breeding Pax8-rtTA/LC1 with NEDD4-2^fl/fl mice (53). The resulting offspring were interbred to obtain NEDD4-2^fl/fl/Pax8-rtTA/LC1 mice. For both SPAK/OSR1 DKO and NEDD4-2 KO mice, Cre-mediated recombination was induced by administration of doxycycline hyclate (5 mg/ml, 5% sucrose) in the drinking water for 2–3 wk. After doxycycline treatment, mice were returned to regular drinking water for ≥2 wk before experiments were performed. Littermate mice of the same age and genotype provided with 5% sucrose in drinking water were used as controls. Mice aged 2–5 mo were used, and both male and female mice were used as indicated in the figures.

PCR genotyping.

Genomic DNA extracts were prepared from tail snips by heating overnight at 55°C in 300 μl digestion solution containing 5 mM EDTA, 200 mM NaCl, 100 mM Tris (pH 8.0), 0.2% SDS, and 0.4 mg/ml proteinase K followed by ethanol precipitation. The following primers were used: Pax8, forward 5′-CCATGTCTAGACTGGACAAGA-3′ and reverse 5′-CAGAAAGTCTTGCCATGACT-3′; Cre, forward 5′-TTTCCCGCAGAACCTGAACCTGAAGAT-3′ and reverse 5′-TCACCGGCATCAACGTTTTCTT-3′; SPAK, forward 5′-TTCTTAGCCACAGGGGGTGATG-3′ and reverse 5′-GAGTCATAGAAGAGCAGAATAGCAG-3′; OSR1flox, forward 5′-AGCTCAGGCTCCCTCCACGGAG-3′ and reverse 5′-AAGACACATTGATACTCTGTTT-3′; NEDD4-2, forward 5′-TGAGCTCATTGCTTCACTTCC-3′ and reverse 5′-TTCATGCTCGAAGCCTTAGC-3′, and NEDD4-2 flox, reverse 5′-TTTGTGAGGACAGCCTCTAGC-3′.

RNA extraction and quantitative PCR.

Snap-frozen kidneys were ground in a mortar and pestle under liquid nitrogen before RNA extraction with RNA universal mini kit (catalog no. 73404, Qiagen). Powdered kidney tissue was dissolved in Qiazol and further mechanically triturateed through a narrow syringe until the solution became homogenous before the manufacturer’s protocol for total RNA extraction was followed. RNA was then quantified on a spectrophotometer, and 1,000 ng of total RNA were converted to cDNA using the Maxima H Minus First Strand cDNA Synthesis Kit (catalog no. K1681, ThermoFischer Scientific) according to the manufacturer’s instructions. The resulting reaction was then diluted 1:10 so that 5 ng template cDNA was input into each quantitative PCR. Predesigned TaqMan probes were purchased from ABI for 18S rRNA and Slc12a3 (NCC). Duplicate quantitative PCRs were quantified on a QuantStudio 7 Flex (ABI). Expression levels for Slc12a3 were normalized to the levels of 18S rRNA and then normalized to the mean expression of Slc12a3 for mice on a control diet.

Antibodies.

Antibody sources, species, dilutions, and references are shown in Table 1.

Table 1.

Antibodies

Antibody (Species-Antigen)	Use	Dilution	Source	Reference(s)
Rabbit-tNCC*	WB	1:6,000	Ellison Laboratory	21, 22, 40, 42
Rabbit-pNCC (T53)*	WB	1:2,000	Ellison Laboratory	21, 22, 40, 42
Rabbit-N-WNK4*	WB	1:1,000	Ellison Laboratory	22, 42
Rabbit-N-WNK4*	IF	1:2,000	Ellison Laboratory	22, 42
Rabbit-C-SPAK*	WB	1:5,000	McCormick Laboratory	21, 40
Rabbit-C-SPAK*	IF	1:5,000	McCormick Laboratory	21, 40
Rabbit-pSPAK (S383)	WB	1:1,000	Millipore (07-2273)
Guinea pig-parvalbumin	IF	1:2,000	Swant, GP72	22
Mouse-calbindin	IF	1:2,000	Swant, calbindin D-28, (300)	22
Rabbit-NEDD4L	WB	1:1,000	Cell Signaling Technology (no. 4013)	53
Rabbit-SGK1	WB	1:1,000	Sigma-Aldrich	26

tNCC, Na⁺-Cl⁻ cotransporter (NCC); pNCC, phosphorylated NCC; WNK4, with-no-lysine (K) kinase-4; STE20/SPS-1-related proline/alanine-rich kinase; pSPAK, phosphorylated SPAK; NEDD4L, neuronal precursor cell developmentally downregulated 4L; SGK1, serum and glucocorticoid-induced kinase 1; WB, Western blot; IF, immunofluorescence.

^*NCC, pNCC, WNK4, SPAK, p-SPAK, and NEDD4L antibodies have been validated in knockout mice.

Kidney Western blot analysis.

Kidneys were snap frozen in liquid nitrogen after being harvested and stored at −80°C until homogenization using a Potter homogenizer in 1 ml cold homogenization buffer containing lysis buffer containing 300 mM sucrose, 50 mM Tris·HCl (pH 7.4), 1 mM EDTA, 1 mM EGTA, 1 mM NaVO₄, 50 mM NaF, 1 mM ditiothreitol (Millipore, Sigma), 1 mM phenylmethane sulfonyl fluoride, 1 µg/ml aprotinin, and 4 µg/ml leupeptin. The homogenate was centrifuged at 6,000 g for 15 min at 4°C, and the supernatant was transferred to a new tube and stored at −80°C. Then, 20 μg protein was separated on a 4–12% bis-Tris acetate gel (Invitrogen) and transferred to a polyvinylidene fluoride membrane using the Trans-Blot Turbo Transfer System (Bio-Rad). The membrane was blocked with 5% nonfat milk in PBS-Tween followed by an incubation with primary antibody for either 1 h at room temperature or overnight at 4°C. Membranes were washed, incubated with horseradish peroxidase-goat anti-rabbit IgG (1:7,500, no. 65-6120, Invitrogen), washed again, and finally incubated with Western Lightning ECL (Perkin-Elmer). The ECL signal was detected with a Syngene Pxi4 imager. For donkey anti-rabbit IgG (1:15,000, LI-COR, IRDye 680RD) secondary antibody, an Odyssey CLx Near-Infrared Fluorescence Imaging System was used to scan the membrane, and densitometry was performed with ImageJ (http://rsbweb.nih.gov/ij/). Protein loading was adjusted by densitometric quantitation of total protein after Coomassie staining (see Supplemental Fig. S1 for an example; Supplemental Data is available online at http://dx.doi.org/10.6084/m9.figshare.8947001) (22, 43).

Blood analysis.

Blood was collected via cardiac puncture under isoflurane anesthesia and transferred into heparinized tubes; 80 μl were loaded into a Chem8+ cartridge for electrolyte measurement by an i-STAT analyzer (Abbot Point of Care). The remainder was centrifuged at 2,000 g for 5 min at room temperature, and plasma was removed and stored at −80°C. Plasma Mg²⁺ concentration was quantified by ratiometric assay (Pointe Science, Canton, MI). For some experiments, plasma K⁺ concentration was determined by flame photometry.

Immunofluorescence.

Animals were injected with an anesthesia cocktail [ketamine-xylazine-acepromazine (50:5:0.5 mg/kg)] and under deep anesthesia were perfusion fixed with 4% paraformaldehyde. After cryoprotection in 800 mosmol sucrose and freezing in OCT, 5-μm sections were cut. Sections were incubated overnight at room temperature with primary antibodies (Table 1). The secondary antibody used was Alexa Fluor 488 goat anti-guinea pig IgG (A11073, Invitrogen, 1:1,000), Cy3 goat anti-rabbit IgG (no. 81-6115, Zymed, 1:1,000), or FITC goat anti-mouse IgG (no. 81-6511, Zymed, 1:1,000). Secondary antibody was incubated for 1 h at room temperature. Images were captured with a Zeiss AXIO Imager M2 fluorescent microscope.

Statistics.

The null hypothesis was tested using two-tailed unpaired t-tests, one-way ANOVA, or two-way ANOVA for repeated measures using GraphPad Prism 7 as indicated in the figures. Post hoc analysis was performed using the Tukey multiple-comparison test. All data are plotted as means ± SE. P < 0.05 was considered significant.

Study approval.

Animal studies were performed in adherence to the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals (National Academies Press, 2011) and approved by the Oregon Health and Science University Institutional Animal Care and Use Committee (protocol IP00286) and New York Medical College Institutional Animal Care and Use Committee (protocol 60-8-1016H).

RESULTS

Long-term (14 days) dietary Mg²⁺ restriction downregulates NCC.

Mice were maintained on normal Mg²⁺, Mg²⁺-deficient, or high-Mg²⁺ diet for 14 days. As expected, plasma Mg²⁺ was significantly lower on the Mg²⁺-deficient diet compared with the normal diet and high-Mg²⁺ diet (Fig. 1A), but both the normal and high-Mg²⁺ diet groups had similar plasma Mg²⁺ concentration (Fig. 1A). In humans, maintenance of plasma Mg²⁺ concentration under a high Mg²⁺ load does not occur via increased renal excretion but by reduced gut absorption (30). To determine whether the high-Mg²⁺ diet increased urinary Mg²⁺ excretion to maintain plasma Mg²⁺ concentration in mice, we measured urinary Mg²⁺ excretion at baseline and after 1, 2, 3, or 14 days of normal or high-Mg²⁺ diet. We found no differences in urinary Mg²⁺ excretion between the normal diet and high-Mg²⁺ diet at any time point (Supplemental Fig. S2). Mg²⁺ inhibits renal outer medullary K⁺ (ROMK) channel-mediated K⁺ activity in vitro (38, 70, 80), so lower plasma Mg²⁺ concentration might be expected to lower plasma K⁺ concentration. However, plasma K⁺ concentration did not differ between the groups (Fig. 1A). Hypokalemia is not typically observed in isolated Mg²⁺ deficiency in experimental models (29, 73), for two possible reasons. First, chronic Mg²⁺ deficiency impairs Na⁺-K⁺-ATPase pump activity, which decreases K⁺ uptake by the muscle and kidney and thus maintains plasma K⁺ concentration (9, 29, 78). Moreover, increased K⁺ secretion through ROMK channels would cause hyperpolarization of the plasma membrane reducing K⁺ secretion through ROMK channels (24, 29). Plasma ionized Ca²⁺ concentration was lower in the Mg²⁺-deficient group than normal diet and high-Mg²⁺ diet groups (Fig. 1A), since chronic Mg²⁺ deficiency impairs the release of parathyroid hormone, leading to lower plasma ionized Ca²⁺ concentration (4, 77).

Fig. 1.

Long-term (14 days) dietary Mg²⁺ restriction downregulates total Na⁺-Cl⁻ cotransporter (tNCC) abundance. A: plasma [K⁺] did not differ among the study groups (one-way ANOVA with Tukey’s multiple-comparison test). Plasma ionized [Ca²⁺] (i-[Ca²⁺]) was significantly lower on a Mg²⁺-deficient diet (D) compared with normal diet (N) and high-Mg²⁺ diet (H) (one-way ANOVA with Tukey’s multiple-comparison test). Plasma [Mg²⁺] was significantly lower on a Mg²⁺-deficient diet compared with normal diet and high-Mg²⁺ diet (one-way ANOVA with Tukey’s multiple-comparison test). B: Western blot analysis of whole kidney lysates showed that tNCC abundance was significantly lower on Mg²⁺-deficient diet than on normal diet and high-Mg²⁺ diet (one-way ANOVA with Tukey’s multiple-comparison test). C: phosphorylated (p)NCC abundance was also significantly lower on Mg²⁺-deficient diet compared with normal and high-Mg²⁺ diets (one-way ANOVA with Tukey’s multiple-comparison test). D: Western blot analysis of whole kidney lysates from female mice showed that tNCC abundance was significantly reduced on Mg²⁺-deficient diet than on normal diet (two-tailed unpaired t-test). For blot quantification, densitometric values were normalized using Coomassie-stained gels. Values are means ± SE; values in parentheses indicate n values. *P < 0.05; **P < 0.001.

Although plasma K⁺ concentration, a strong regulator of NCC, did not differ between groups (Fig. 1A), the abundance of total NCC in the kidney was dramatically lower in mice maintained on a Mg²⁺-deficient diet compared with those on normal or high-Mg²⁺ diets (Fig. 1B). The abundance of pNCC was also significantly lower in mice maintained on a Mg²⁺-deficient diet (Fig. 1C). However, the pNCC-to-total NCC ratio did not differ between groups, suggesting that the primary effect is on total NCC abundance. Although female mice display a greater abundance of total NCC and pNCC on a normal diet (52), dietary Mg²⁺ restriction also resulted in lower total NCC abundance in female mice (Fig. 1D), indicating no sex difference. Considering that no differences were observed in total NCC or pNCC between normal and high-Mg²⁺ diets, subsequent studies compared only the effects of normal and low-Mg²⁺ diets.

Dietary Mg²⁺ restriction does not affect the WNK4/SPAK pathway.

The stability of NCC at the plasma membrane and its overall abundance may depend on its phosphorylation status (54). To determine whether dietary Mg²⁺ deficiency altered activity of the canonical NCC-activating pathway, we first assessed the abundance of WNK4, the upstream kinase in the pathway. WNK4 abundance did not differ between normal and low-Mg²⁺ diets (Fig. 2A). SPAK and OSR1 lie downstream of WNK4 and directly phosphorylate the NH₂-terminus of NCC (50). Abundances of pSPAK and total SPAK did not differ between normal and Mg²⁺-deficient diets (Fig. 2B). These data suggest that the WNK-SPAK/OSR1 pathway is neither activated nor inhibited by dietary Mg²⁺ restriction, but to further test this, we performed immunofluorescence to determine whether WNK4 and SPAK cellular localization changed. SPAK and WNK4 have been shown to localize to intracellular puncta (“WNK bodies”) in response to dietary K⁺ restriction (8, 22, 62). The composition and function of these puncta are not well defined, and it is unknown whether their formation plays a regulatory role in NCC and DCT function. Immunofluorescence showed that dietary Mg²⁺ restriction did not cause any change in the intracellular distribution of SPAK or WNK4 along DCT1, as shown by colocalization with parvalbumin (Fig. 2, C and D) or the DCT2/connecting segment (CNT), as shown by colocalization with calbindin (Supplemental Fig. S3).

Fig. 2.

Long-term (14 days) dietary Mg²⁺ restriction did not affect with-no-lysine (K) kinase-4 (WNK4) or STE20/SPS-1-related proline/alanine-rich kinase (SPAK). A: Western blot analysis of whole kidney lysates did not show any significant difference in the abundance of WNK4 between normal diet (N) and Mg²⁺-deficient (D) diet groups (two-tailed unpaired t-test). B: full-length SPAK (FL-SPAK) and phosphorylated SPAK (pSPAK) abundances also did not differ between normal and Mg²⁺-deficient diets (two-tailed unpaired t-test). C: immunofluorescence for SPAK (red) showed similar localization in normal and Mg²⁺-deficient diets. Colocalization with parvalbumin (PV; green) showed that SPAK was localized along the distal convoluted tubule (DCT). D: WNK4 (red) localization did not differ between normal diet and Mg²⁺-deficient diet. PV (green) was used to identify DCT in the same section. Note that on normal diet and low-Mg²⁺ diet, WNK4 signal is weak and diffuse along the DCT and DCT2/connecting segment (CNT), in contrast to what is observed after low-K⁺ diet (62) or genetic activation of the WNK4-SPAK pathway (22), so the images were not overlaid. Results are representative of 5 independent experiments. Scale bars in C and D = 40 μm. E: dietary Mg²⁺ restriction downregulated total Na⁺-Cl⁻ cotransporter (tNCC) abundance in SPAK knockout (KO) mice. tNCC abundance was also significantly reduced in SPAK and oxidative stress-response kinase-1 (OSR1) double-KO (DKO) mice on Mg²⁺-deficient diet compared with normal diet (two-tailed unpaired t-test). For blot quantification, densitometric values were normalized using Coomassie-stained gels. Values are means ± SE; values in parentheses indicate n values. *P < 0.01; **P < 0.001.

Our analysis of SPAK and WNK4 abundance and distribution suggested that SPAK and WNK4 are not directly affected by dietary Mg²⁺ restriction and that altered activity of the pathway does not play a role in NCC downregulation. To further confirm this, we used SPAK KO and SPAK/OSR1 DKO mice. SPAK is the major NCC activator in vivo, since its disruption in mice leads to a dramatic reduction (∼90%) in pNCC and total NCC (27, 40, 82). SPAK KO mice were maintained on normal or Mg²⁺-deficient diets for 14 days. Dietary Mg²⁺ restriction further decreased total NCC abundance in SPAK KO mice (Fig. 2E). We have previously reported that OSR1 can compensate for a lack of SPAK to activate NCC in response to dietary K⁺ restriction (21). To exclude a compensatory role for OSR1, SPAK/OSR1 DKO mice were placed on a normal or Mg²⁺-deficient diet. After 14 days of dietary Mg²⁺ restriction, total NCC abundance was markedly lower in DKO mice (Fig. 2E). Seeing that total NCC abundance was lower after dietary Mg²⁺ restriction in the presence or absence of SPAK/OSR1, our data suggest that altered activity of the WNK-SPAK/OSR1 pathway is not necessary for the effect of low plasma Mg²⁺ concentration on NCC.

Short-term dietary Mg²⁺ restriction does not change plasma K⁺ concentration despite lower NCC abundance.

A previous study (66) examining the effect of dietary Mg²⁺ on NCC involved chronic manipulation (14 days). In contrast, studies examining the effects of manipulating K⁺ or NaCl have involved much shorter durations [typically 3–7 days for low K⁺ (21, 62, 71), 15 min to 4–7 days for high K⁺ (10, 12, 58, 65), or 7–10 days for low or high NaCl (12, 65)]. Furthermore, Ortega et al. (44) reported that short-term (3 days) dietary Mg²⁺ deficiency significantly increased plasma K⁺ concentration. Acute elevation of plasma K⁺ concentration after gavage of a KCl bolus caused rapid dephosphorylation of NCC (47, 58), which could affect total NCC abundance (54). Thus, one possibility is that in the short term, lower plasma Mg²⁺ concentration decreases total NCC abundance secondarily to an increase in plasma K⁺ concentration. Wild-type (WT) mice were placed on a normal or Mg²⁺-deficient diet for 3 days. In contrast to the previous report by Ortega et al. (44), plasma K⁺ concentration did not differ between normal and Mg²⁺-deficient groups (Fig. 3A), but total NCC and pNCC abundances were significantly lower in the Mg²⁺-deficient group (Fig. 3, B and C). No significant difference was observed in the abundance of total NCC in mice maintained on normal or high-Mg²⁺ diets for 3 days (data not shown). Because NCC abundance was lower after Mg²⁺ restriction, we measured 24-h urinary Na⁺ and K⁺ excretion. Short-term restriction did not alter body weight [26.6 ± 0.6 normal (n =6) vs. 27.4 ± 0.3 g low Mg²⁺ (n = 5), mean ± SE, P = 0.28], but food intake was lower (3.99 ± 0.16 vs. 3.36 ± 0.12 g, P = 0.01). However, normalized Na⁺ and K⁺ excretion did not differ (Na⁺: 3.29 ± 0.15 vs. 3.42 ± 0.11 μmol·g body wt⁻¹·g food intake⁻¹, P = 0.52; K⁺: 6.58 ± 0.32 vs. 6.36 ± 0.27 μmol·g body wt⁻¹·g food intake⁻¹, P = 0.63). With long-term Mg²⁺ restriction, there were no differences in body weight, food intake, or 24-h urinary Na⁺ or K⁺ excretion (Supplemental Fig. S4). These data are similar to those of Van Angelen et al. (66), who observed no difference in 24-h urinary Na⁺ and K⁺ excretion between mice on a Mg²⁺-deficient diet or high-Mg²⁺ diet after 14 days of restriction. With respect to functional effects on NCC, it should be noted that mice heterozygous for a Gitelman syndrome mutation (with 50% lower functional NCC) did not display changes in plasma or urinary electrolyte excretion at baseline (83).

Fig. 3.

Short-term (3 days) dietary Mg²⁺ restriction downregulates total Na⁺-Cl⁻ cotransporter (tNCC). A: plasma [K⁺] did not differ between the diets (two-tailed unpaired t-test). Plasma ionized [Ca²⁺] (i-[Ca²⁺]) was significantly lower on Mg²⁺-deficient diet (D) compared with normal diet (N) (two-tailed unpaired t-test), and plasma [Mg²⁺] was significantly lower on Mg²⁺-deficient diet than on normal diet (two-tailed unpaired t-test) B: Western blot analysis of whole kidney lysates showed that tNCC abundance was significantly lower on Mg²⁺-deficient diet than on normal diet (two-tailed unpaired t-test). C: abundance of phosphorylated (p)NCC was also significantly lower on Mg²⁺-deficient diet compared with normal diet (two-tailed unpaired t-test). For blot quantification, densitometric values were normalized using Coomassie-stained gels. Values are means ± SE; values in parentheses indicate n values. *P < 0.05; **P = 0.0001.

NEDD4-2 participates in NCC downregulation.

NEDD4-2 ubiquitinates NCC and the epithelial Na⁺ channel (ENaC), tagging them for proteasomal degradation (18, 53). The precise target of NEDD4-2 varies in response to different dietary manipulations, with NCC preferred during NaCl loading and ENaC preferred during dietary K⁺ restriction (1, 51). To determine whether NEDD4-2 plays a role in the dietary Mg²⁺ restriction-mediated downregulation of NCC, inducible nephron-specific NEDD4-2 KO mice were maintained on normal diet and Mg²⁺-deficient diet for 3 days. As previously reported, tNCC and pNCC abundances were higher in NEDD4-2-KO mice compared with control mice (53). However, in contrast to our observations in WT mice, dietary Mg²⁺ restriction did not lower the abundance of total NCC in NEDD4-2 KO mice (Fig. 4A). These data suggest that NEDD4-2 participates in the downregulation of NCC during Mg²⁺ deficiency. In this model, NEDD4-2 is knocked out in tubule segments but not in nontubule cells (53). Dietary Mg²⁺ restriction appeared to lead to higher abundance of nontubule NEDD4-2 in KO mice. Quantification of NEDD4-2 abundance and analysis by two-way ANOVA revealed no effect of dietary Mg²⁺ restriction on NEDD4-2 abundance in control mice [100 ± 9 normal (n = 3) vs. 105 ± 18% low Mg²⁺ (n = 3), mean ± SE, P = 0.98,]. NEDD4-2 abundance was significantly lower in KO mice versus control mice, but there were no differences in NEDD4-2 abundance between KO mice on a normal or low-Mg²⁺ diet (P = 0.47). However, comparison with control mice (with high NEDD4-2 abundance) likely masks an effect, so we directly compared KO mice on normal and low Mg²⁺ diets by an unpaired t-test and found significantly higher nontubule NEDD4-2 abundance after Mg²⁺ restriction [100 ± 13 normal (n = 3) vs. 401 ± 62% low Mg²⁺ (n = 3), mean ± SE, P = 0.008 ]. These data suggest that Mg²⁺ restriction exerts effects in nontubular cells in the kidney, which might also lead to the significantly higher abundance of total NCC in NEDD4-2 KO mice on a low-Mg²⁺ diet compared with normal diet (Fig. 4A).

Fig. 4.

Disruption of tubular E3 ubiquitin-protein ligase. Neuronal precursor cell developmentally downregulated 2 (NEDD4-2) prevents dietary Mg²⁺ restriction-mediated downregulation of Na⁺-Cl⁻ cotransporter (NCC). A: control mice had lower abundance of total NCC (tNCC) and phosphorylated (p)NCC on Mg²⁺-deficient diet (D) compared with normal diet (N), whereas in inducible nephron-specific NEDD4-2 knockout (KO) mice, dietary Mg²⁺ restriction did not lead to lower abundance of NCC (two-way ANOVA with Tukey’s multiple-comparison test). B: abundance of serum and glucocorticoid-induced kinase 1 (SGK1) was comparable between normal diet and Mg²⁺-deficient diet after 3 days of dietary Mg²⁺ manipulation (two-tailed unpaired t-test). C: NCC mRNA did not differ among normal diet or short-term (3 days) and long-term (14 days) dietary Mg²⁺ restriction (one-way ANOVA with Tukey’s multiple-comparison test). For blot quantification, densitometric values were normalized using Coomassie-stained gels. Values are means ± SE; values in parentheses indicate n values. *P < 0.05; **P < 0.01; ***P < 0.0001.

Expression of serum and glucocorticoid-induced kinase 1 (SGK1) is strongly induced by aldosterone and phosphorylates and inhibits NEDD4-2 to increase ENaC abundance at the plasma membrane (33). There were no differences in SGK1 abundance after Mg²⁺ restriction (Fig. 4B), suggesting that the NEDD4-2 regulatory pathway differs between Na⁺ and Mg²⁺ restriction. It has been previously reported that NCC mRNA does not differ between mice fed low- or high-Mg²⁺ diets for 14 days (14, 66). Similarly, we found no significant differences in NCC mRNA between mice on low-Mg²⁺ or normal Mg²⁺ diets in the short term or long term (Fig. 4C), suggesting posttranscriptional regulation of NCC abundance.

Dietary Mg²⁺ restriction overrides the stimulatory effect of dietary K⁺ restriction on NCC.

Because dietary K⁺ restriction is a potent activator of NCC, we next determined whether Mg²⁺ deficiency affects K⁺ deficiency-mediated activation of NCC. WT mice were placed on K⁺-deficient, Mg²⁺-deficient, or Mg²⁺-K⁺-deficient diets for 3 days. Plasma Mg²⁺ concentration was markedly lower on Mg²⁺-deficient and Mg²⁺-K⁺-deficient diets compared with K⁺-deficient diet, and plasma K⁺ concentration was significantly decreased on K⁺-deficient and Mg²⁺-K⁺-deficient diets compared with Mg²⁺-deficient diet (Fig. 5A). There were no synergistic effects of dietary Mg²⁺-K⁺ restriction with regard to plasma Mg²⁺ or K⁺ concentrations. The abundances of total NCC and pNCC were lower on Mg²⁺-deficient and Mg²⁺-K⁺-deficient diets than on K⁺-deficient diet (Fig. 5, B and C), suggesting that dietary Mg²⁺ deficiency overrode the ability of dietary K⁺ restriction to induce NCC phosphorylation. To determine whether this effect persisted over the long term, WT mice were maintained on normal, Mg²⁺-K⁺-deficient, Mg²⁺-deficient, or K⁺-deficient diets for 14 days. Dietary Mg²⁺, K⁺, or Mg²⁺-K⁺ restriction resulted in lower plasma Mg²⁺ concentration and/or K⁺ concentration, with no synergistic effects (Fig. 6A). After 14 days of Mg²⁺ restriction, total NCC and pNCC abundances were significantly lower compared with control diet (Fig. 6B), consistent with Fig. 1B. Total NCC and pNCC abundances were also significantly lower after Mg²⁺-K⁺-deficient diet compared with control diet (Fig. 6B). As observed after short-term restriction (Fig. 5, B and C), abundances of total NCC and pNCC did not differ between mice placed on Mg²⁺-deficient or Mg²⁺-K⁺-deficient diets (Fig. 6B). Unexpectedly, given the well-established effect of short-term K⁺ restriction to strongly increase pNCC abundance, the abundances of total NCC and pNCC on the K⁺-deficient diet compared with the normal diet did not differ (Fig. 6B).

Fig. 5.

Short-term (3 days) dietary Mg²⁺ restriction overrides the effect of K⁺ deficiency on Na⁺-Cl⁻ cotransporter (NCC). A: plasma [K⁺] was significantly lower on K⁺-deficient (D) diet and on Mg²⁺-K⁺ deficient diet compared with normal diet (N) and Mg²⁺-deficient diet (one-way ANOVA with Tukey’s multiple-comparison test). Plasma [Mg²⁺] was significantly lower on Mg²⁺-deficient and Mg²⁺-K⁺-deficient diet compared with normal diet and K⁺-deficient diet (one-way ANOVA with Tukey’s multiple-comparison test). B: Western blot analysis of whole kidney lysates showed that total NCC (tNCC) abundance was significantly lower on Mg²⁺-deficient diet and on Mg²⁺-K⁺-deficient diet than on normal diet and K⁺-deficient diet (one-way ANOVA with Tukey’s multiple-comparison test). C: phosphorylated (p)NCC abundance was significantly higher on K⁺-deficient diet than all other study groups (one-way ANOVA with Tukey’s multiple-comparison test). For blot quantification, densitometric values were normalized using Coomassie-stained gels. Values are means ± SE; values in parentheses indicate n values. *P < 0.01; **P < 0.001; ***P < 0.0001.

Fig. 6.

Compared with normal diet, long-term (14 days) dietary K⁺ restriction did not increase phosphorylated Na⁺-Cl⁻ cotransporter (pNCC) abundance. A: plasma [K⁺] was significantly lower on Mg²⁺-K⁺-deficient diet (D) and on K⁺-deficient diet compared with normal diet (N) and Mg²⁺-deficient diet (one-way ANOVA with Tukey’s multiple-comparison test). Plasma [Mg²⁺] was significantly lower on Mg²⁺-K⁺-deficient diet and on Mg²⁺-deficient diet compared with normal diet and K⁺-deficient diet (one-way ANOVA, Tukey’s multiple-comparison test). B: Western blot analysis of whole kidney lysates showed that long-term (14 days) dietary Mg²⁺ restriction downregulated total Na⁺-Cl⁻ cotransporter (tNCC) abundance compared with normal diet and K⁺-deficient diet (one-way ANOVA with Tukey’s multiple-comparison test). Long-term (14 days) dietary Mg²⁺ restriction also lowered pNCC abundance compared with normal diet, whereas pNCC abundance was not higher on K⁺-deficient diet (one-way ANOVA with Tukey’s multiple-comparison test). For blot quantification, densitometric values were normalized using Coomassie-stained gels. Values are means ± SE; values in parentheses indicate n values. *P < 0.05; **P < 0.001; ***P < 0.0001.

DISCUSSION

The present study reveals that both long-term and short-term dietary Mg²⁺ deficiency downregulate total NCC abundance. Our observation of downregulation of total NCC protein abundance by dietary Mg²⁺ deficiency confirms previous findings (66) using Western blot analysis, rather than immunofluorescence, which is more difficult to accurately quantitate. In addition to using high- and low-Mg²⁺ diets, we also used a diet with a level of Mg²⁺ similar to that of regular chow and found that high Mg²⁺ did not affect NCC abundance. It was recently reported that mice expressing constitutively active SPAK, which display hyperactive NCC and mimic the disease familial hyperkalemic hypertension, are not hypermagnesemic (69). Therefore, NCC activity does not appear to be important for mediating long-term Mg²⁺ elimination during high Mg²⁺ intake. Our data also provide mechanistic insights. The effect of dietary Mg²⁺ restriction does not involve the WNK/SPAK/OSR1 pathway, the major NCC-activating pathway, but the E3 ubiquitin-protein ligase NEDD4-2 is required for Mg²⁺ deficiency-induced downregulation of NCC. We also found that NCC abundance and phosphorylation are dramatically lower after combined Mg²⁺-K⁺ restriction, despite the well-established effect of short-term K⁺ restriction to strongly activate NCC. Finally, we made the unexpected observation that total NCC and pNCC abundances are not higher than those in controls after long-term K⁺ restriction.

It is well established that Mg²⁺ restriction impairs the release of parathyroid hormone, leading to lower plasma ionized Ca²⁺ concentration, which might contribute to downregulation of NCC abundance. Using [³H]metolazone binding, Fanestil et al. (19) reported no effect of dietary Ca²⁺ restriction or loading on NCC density in the rat kidney. However, dietary Ca²⁺ manipulation had no effect on plasma ionized Ca²⁺ concentration. Supplementation of dietary Ca²⁺ to maintain plasma ionized Ca²⁺ concentration during Mg²⁺ restriction might help address this question.

Ortega et al. (44) reported that serum K⁺ concentration was significantly higher after short-term (3 days) dietary Mg²⁺ deficiency. Plasma K⁺ concentration potently regulates the abundance and phosphorylation status of NCC (21, 61, 62). High dietary K⁺ intake causes rapid dephosphorylation and, hence, inactivation of NCC (47, 58), which could contribute to degradation of NCC (54). However, in our hands, short-term dietary Mg²⁺ deficiency did not lead to a difference in plasma K⁺ concentration. This could be due to differences in technique (both studies used cardiac puncture) during the blood draw procedure, which may cause hemolysis. Ortega et al. (44) reported an average serum K⁺ concentration of ∼5 mmol/l at baseline, whereas in WT mice maintained on a normal diet, we typically observed a plasma K⁺ concentration of 3.5−4.5 mmol/l (21, 22). Moreover, our observation is consistent with the reports of other studies that in experimental models isolated Mg²⁺ deficiency does not cause any significant change in plasma K⁺ concentration (29, 73). Therefore, we conclude that Mg²⁺ deficiency does not lower total NCC abundance through an effect on plasma K⁺ concentration.

Unexpectedly, we did not observe increased abundances of total NCC and pNCC after 14 days of a K⁺-deficient diet. We repeated the experiment with a second batch of mice and obtained the same results (data not shown). Recent studies examining the effects of dietary K⁺ restriction on NCC have been performed over a period of 3−7 days (21, 62, 65, 71). A study from the early 1990s (17) has reported dramatic changes in tubule structure in rat kidney after long-term (18 days) dietary K⁺ restriction. Compared with controls, kidney weight was increased by 30% and glomerular filtration rate (GFR) decreased by 14% on a long-term K⁺-deficient diet. Striking hypertrophic changes were reported to occur in the outer medullary collecting duct (17). We recently reported that short-term (3 days) K⁺ restriction induces proliferation in early DCT and aquaporin 2-positive cells but does not determine the effects of sustained K⁺ restriction on proliferation or apoptosis in these segments (56). Thus, one possibility is that changes in NCC activity and DCT hypertrophy contribute to short/medium-term responses to K⁺ restriction, whereas remodeling of the K⁺-secreting CNT and cortical collecting duct becomes the predominant mechanism over the long term. Further studies will be required to further explore the precise changes in tubule structure at the molecular level after long-term K⁺ restriction and also whether Mg²⁺ restriction induces tubule remodeling and modulates K⁺ restriction-induced changes.

Short- or long-term Mg²⁺ restriction did not affect NCC mRNA levels, consistent with previous reports using only Mg²⁺-deficient and high-Mg²⁺ diets (14, 66). Therefore, we hypothesized that the underlying cause of downregulation of NCC after dietary Mg²⁺ restriction involves an effect on NCC protein degradation. The ubiquitin protein ligase NEDD4-2 is a well-known regulator of NCC abundance (18, 53). The primary regulatory target of NEDD4-2 action changes depending on dietary manipulation. NEDD4-2 primarily regulates NCC on dietary Na⁺ loading, whereas on dietary K⁺ restriction it primarily regulates ENaC (1, 51). We found that NCC abundance was not lower after Mg²⁺ restriction in NEDD4-2 KO mice, suggesting that NEDD4-2 may mediate the effect. NEDD4-2 can be activated by either directly binding Ca²⁺ via its Calb/C2-binding domain, which releases its autoinhibition, or after binding to substrate (31, 72). In the kidney, 100% of intracellular Mg²⁺ can be exchanged with the plasma within 3 to 4 h (36). Therefore, both short-term and long-term dietary Mg²⁺ restriction would be expected to significantly reduce intracellular Mg²⁺ content. Because intracellular Mg²⁺ concentration inversely regulates intracellular free ionized Ca²⁺ concentration (25, 84), dietary Mg²⁺ restriction likely elevates intracellular free ionized Ca²⁺ concentration. We speculate that along the DCT this might activate NEDD4-2, leading to NCC degradation. However, one caveat is that NEDD4-2 might prevent degradation of NCC in any condition. Mg²⁺ deficiency is also known to impair Na⁺-K⁺-ATPase pump activity (23, 34). Impaired Na⁺-K⁺-ATPase pump function diminishes NCC activity, and inactive NCC is more susceptible to degradation (34, 39, 54, 60, 81). Reduced K⁺ concentration also induces a K⁺-dependent decrease in Na⁺-K⁺-ATPase pump activity, which has been proffered to cause chronic dietary K⁺ restriction-mediated renal hypertrophy (7, 63). Further studies are required to determine the cellular mechanisms by which dietary Mg²⁺ restriction mediates NCC degradation.

There are limitations to our study. First, Mg²⁺ deficiency has effects on the vasculature that have been proposed to cause increased blood pressure (32), although it took 8 wk of Mg²⁺ restriction to cause a blood pressure increase. Downregulation of NCC could act as a compensatory mechanism to maintain blood pressure during the early phase of Mg²⁺ restriction. This could be tested by administration of the vasodilator hydralazine during Mg²⁺ restriction. Next, the recommended daily required Mg²⁺ intake for mice is inconsistent, but with a range of 500−2,600 mg/kg (43a). Whereas our normal diet contained 1,500 mg/kg and our high-Mg²⁺ diet contained 4,800 mg/kg, our low-Mg²⁺ diet provided only a maximum of 30 mg/kg, which is extremely low and nonphysiological. However, this low-Mg²⁺ diet resulted in plasma Mg²⁺ concentration of ∼0.4 mmol/l, which is in the range of values that have been observed with alcoholism, furosemide treatment, short bowel and other malabsorptive disorders, or rare diseases such as familial hypomagnesemia with hypercalciuria and nephrocalcinosis caused by claudin 16 or claudin 19 mutations. Nevertheless, future studies will require titration of dietary Mg²⁺ to minimize effects beyond NCC downregulation. Finally, we found that 14 days on a high-Mg²⁺ diet did not alter NCC abundance, but at the same time, this diet did not increase plasma Mg²⁺ concentration. Our findings are consistent with a previous report (66), and our data suggest that this may be due to lack of gut absorption, since urinary Mg²⁺ excretion did not differ compared with normal diet. However, a recent report by Liu et al. (35) showed that 6 wk of dietary Mg²⁺ loading increased plasma Mg²⁺ concentration by 61%. Together, these observations suggest that over time the mechanisms allowing the gut to limit Mg²⁺ absorption during a high-Mg²⁺ load may break down. Therefore, additional studies are required to determine whether hypermagnesemia induced by more chronic dietary Mg²⁺ loading alters NCC abundance.

In clinical cases of combined hypokalemia and hypomagnesemia, it is essential to coadminister Mg²⁺ to correct the hypokalemia (29). Huang and Kuo (29) proposed that increased Na⁺ delivery to the CNT and cortical collecting duct is a key stimulator of K⁺ wasting during hypomagnesemia. Downregulation of tNCC due to hypomagnesemia might be one mechanism leading to increased distal Na⁺ delivery, maintaining hypokalemia (45, 46, 60). However, isolated hypokalemia potently activates NCC (21, 62), which would be predicted to counter the effect of hypomagnesemia. Our data reveal that dietary Mg²⁺ can override the activating effect of dietary K⁺ restriction on NCC, which could sustain distal Na⁺ delivery and, hence, kaliuresis (Fig. 7). This provides a plausible explanation for the requirement to correct concomitant hypomagnesemia to normalize hypokalemia in the clinical setting.

Fig. 7.

Model of the effect of dietary Mg²⁺ deficiency on total Na⁺-Cl⁻ cotransporter (tNCC) in the setting of hypokalemia. With-no-lysine (K) kinase-4 (WNK4)-STE20/SPS-1-related proline/alanine-rich kinase (SPAK) is the major signaling pathway to activate NCC, whereas E3 ubiquitin-protein ligase neuronal precursor cell developmentally downregulated 4-2 (NEDD4-2) plays a key role in the catabolism of NCC. It is well established that dietary K⁺ deficiency activates NCC by phosphorylation through the WNK4-SPAK signaling pathway along the distal convoluted tubule. During K⁺ deficiency, NEDD4-2 spares NCC and mainly targets the epithelial Na⁺ channel. Dietary Mg²⁺ deficiency downregulates NCC, possibly through NEDD4-2. In the state of both Mg²⁺ and K⁺ deficiency, downregulation of NCC abundance, due to Mg²⁺ deficiency, prevents K⁺ deficiency-mediated NCC activation by phosphorylation through WNK4-SPAK. The effect of Mg²⁺ deficiency on NCC is independent of WNK4 and SPAK.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-098141 (to J. A. McCormick) and DK-54983 and DK-115366 (to D. Lin), American Heart Association Postdoctoral Fellowship 17POST33670206 (to M. Z. Ferdaus), and Swiss National Science Foundation Grant No. 310030_159735 and a grant from the Leducq Foundation (to O. Staub).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

M.Z.F., L.N.M., A.T., and J.A.M. conceived and designed research; M.Z.F., A.M., J.W.N., P.J.B., L.N.M., A.T., O.S., D.-H.L., and J.A.M. performed experiments; M.Z.F., A.M., J.W.N., P.J.B., L.N.M., A.T., O.S., D.-H.L., and J.A.M. analyzed data; M.Z.F., P.J.B., L.N.M., A.T., and J.A.M. interpreted results of experiments; M.Z.F., A.M., L.N.M., and D.-H.L. prepared figures; M.Z.F., L.N.M., and J.A.M. drafted manuscript; M.Z.F., J.W.N., P.J.B., L.N.M., O.S., D.-H.L., and J.A.M. edited and revised manuscript; M.Z.F., A.M., J.W.N., P.J.B., L.N.M., A.T., O.S., D.-H.L., and J.A.M. approved final version of manuscript.