Endothelin‐1 mediates natriuresis but not polyuria during vitamin D‐induced acute hypercalcaemia

Abstract

Key points

• Hypercalcaemia can occur under various pathological conditions, such as primary hyperparathyroidism, malignancy or granulomatosis, and it induces natriuresis and polyuria in various species via an unknown mechanism.

• A previous study demonstrated that hypercalcaemia induced by vitamin D in rats increased endothelin (ET)‐1 expression in the distal nephron, which suggests the involvement of the ET system in hypercalcaemia‐induced effects.

• In the present study, we demonstrate that, during vitamin D‐induced hypercalcaemia, the activation of ET system by increased ET‐1 is responsible for natriuresis but not for polyuria.

• Vitamin D‐treated hypercalcaemic mice showed a blunted response to amiloride, suggesting that epithelial sodium channel function is inhibited.

• We have identified an original pathway that specifically mediates the effects of vitamin D‐induced hypercalcaemia on sodium handling in the distal nephron without affecting water handling.

Abstract

Acute hypercalcaemia increases urinary sodium and water excretion; however, the underlying molecular mechanism remains unclear. Because vitamin D‐induced hypercalcaemia increases the renal expression of endothelin (ET)‐1, we hypothesized that ET‐1 mediates the effects of hypercalcaemia on renal sodium and water handling. Hypercalcaemia was induced in 8‐week‐old, parathyroid hormone‐supplemented, male mice by oral administration of dihydrotachysterol (DHT) for 3 days. DHT‐treated mice became hypercalcaemic and displayed increased urinary water and sodium excretion compared to controls. mRNA levels of ET‐1 and the transcription factors CCAAT‐enhancer binding protein β and δ were specifically increased in the distal convoluted tubule and downstream segments in DHT‐treated mice. To examine the role of the ET system in hypercalcaemia‐induced natriuresis and polyuria, mice were treated with the ET‐1 receptor antagonist macitentan, with or without DHT. Mice treated with both macitentan and DHT displayed hypercalcaemia and polyuria similar to that in mice treated with DHT alone; however, no increase in urinary sodium excretion was observed. To identify the affected sodium transport mechanism, we assessed the response to various diuretics in control and DHT‐treated hypercalcaemic mice. Amiloride, an inhibitor of the epithelial sodium channel (ENaC), increased sodium excretion to a lesser extent in DHT‐treated mice compared to control mice. Mice treated with either macitentan+DHT or macitentan alone had a similar response to amiloride. In summary, vitamin D‐induced hypercalcaemia increases the renal production of ET‐1 and decreases ENaC activity, which is probably responsible for the rise in urinary sodium excretion but not for polyuria.

Key points

• Hypercalcaemia can occur under various pathological conditions, such as primary hyperparathyroidism, malignancy or granulomatosis, and it induces natriuresis and polyuria in various species via an unknown mechanism.

• A previous study demonstrated that hypercalcaemia induced by vitamin D in rats increased endothelin (ET)‐1 expression in the distal nephron, which suggests the involvement of the ET system in hypercalcaemia‐induced effects.

• In the present study, we demonstrate that, during vitamin D‐induced hypercalcaemia, the activation of ET system by increased ET‐1 is responsible for natriuresis but not for polyuria.

• Vitamin D‐treated hypercalcaemic mice showed a blunted response to amiloride, suggesting that epithelial sodium channel function is inhibited.

• We have identified an original pathway that specifically mediates the effects of vitamin D‐induced hypercalcaemia on sodium handling in the distal nephron without affecting water handling.

Abbreviations

AVP

arginine vasopressin

BW

body weight

CaSR

calcium‐sensing receptor

CD

collecting duct

CCD

cortical collecting duct

CEBP

CCAAT/enhancer‐binding protein

CNT

connecting tubule

COPAS

Complex Object Parametric Analyzer and Sorter

DCT

distal convoluted tubule

DHT

dihydrotachysterol

ENaC

epithelial sodium channel

ET

endothelin

ET_A

endothelin type A

ET_B

endothelin type B

HCTZ

hydrochlorothiazide

NDCBE

sodium‐dependent chloride bicarbonate exchanger

NKCC

sodium potassium chloride co‐transporter

NCC

sodium chloride co‐transporter

OMCD

outer medullary collecting duct

PTH

parathyroid hormone

TAL

thick ascending limb

Introduction

Acute hypercalcaemia is one cause of natriuresis and polyuria. Hypercalcaemia can occur under various pathological conditions, such as primary hyperparathyroidism, malignancy or granulomatosis, and its effect has been studied in humans and various animal models, such as mice, rats, dogs and monkeys (Wolf & Ball, 1949; Levitt et al. 1957; DiBona, 1971; Battula et al. 2012). At present, the molecular mechanism underlying the disturbance in water and sodium transport accompanying hypercalcaemia is not well understood. Hypercalcaemia activates the calcium‐sensing receptor (CaSR) in the plasma membrane of many organs, such as parathyroid glands and kidneys (Riccardi & Kemp, 2012). An early effect of CaSR activation is to downregulate the release of parathyroid hormone (PTH) and the synthesis of 1,25(OH)₂‐vitamin D, thereby reducing intestinal absorption and renal reabsorption of calcium. This allows for the elimination of excess calcium from the extracellular fluid (Houillier et al. 2003). However, acute hypercalcaemia‐induced natriuresis and polyuria are, at least in part, independent of PTH and 1,25(OH)₂‐vitamin D (Fuleihan et al. 1998; Adami & Parfitt, 2000).

A vitamin D‐induced hypercalcaemic rat model has been widely used to characterize some of the effects of hypercalcaemia. Hypercalcaemic rats show reduced permeability to sodium in isolated cortical thick ascending limb of the loop of Henle (TAL) (Peterson, 1990) and decreased expression of the sodium potassium chloride co‐transporter NKCC2 (Wang et al. 2002 a). This suggests that hypercalcaemia may induce natriuresis by decreasing NKCC2‐mediated sodium transport in the TAL. Hypercalcaemic rats exhibit a blunted response to arginine vasopressin (AVP) in isolated inner medullary collecting duct (Sands et al. 1998) and reduced expression of the water channel aquaporin 2 (Earm et al. 1998; Wang et al. 2002 b).

To our knowledge, it remains unknown whether other sodium transporters expressed downstream of the cortical TAL, such as the sodium chloride co‐transporter (NCC) in the distal convoluted tubule (DCT) and the epithelial sodium channel (ENaC) in the principal cells of connecting tubules (CNT) and collecting duct (CD), contribute to natriuresis.

Shiraishi et al. (2003) showed increased endothelin (ET)‐1 expression in the distal nephron of vitamin D‐treated hypercalcaemic rats and suggested that it was secondary to the activation of the calcium‐sensing receptor CaSR by hypercalcaemia. ET‐1 can activate both ET type‐A (ET_A) and type‐B (ET_B) receptors (Barton & Yanagisawa, 2008). The renal ET system is important for sodium and water transport (Sandgaard & Bie, 1996). Several in vitro studies have shown an inhibitory effect of ET‐1 on sodium reabsorption in isolated inner medullary collecting duct and cortical collecting duct (CCD) that involves a decrease in Na⁺,K⁺‐ATPase (Zeidel et al. 1989) and ENaC activities, respectively (Bugaj et al. 2008). Moreover, deletion of the ET_B receptor‐encoding gene, Ednrb, in CD cells, or pharmacological inhibition of ET_A and ET_B receptors in isolated CCD has been shown to augment ENaC activity in mice (Bugaj et al. 2012; Lynch et al. 2013). Additionally, inactivation of the ET_A receptor in mouse CD alters the normal response to AVP (Ge et al. 2005). Finally, renal ET‐1 expression is increased on a high Na diet (Fattal et al. 2004).

Methods

Ethical approval

All animal experiments were performed in the accordance of the ethical guideline at the Centre de Recherche des Cordeliers and French legislation of animal care and experimentation (CEEA34.PH.02913).

Animals

Eight‐week‐old C57BL6J male mice were purchased from Charles River (L'Arbresle, France). All mice were maintained under 12:12 h light/dark cycles with access to water ad libitum and standard laboratory diet A04 (SAFE, Augy, France) before the experiments. For clamping the blood PTH level, mice were anaesthesized with isoflurane and osmotic mini pumps (Alzet 1002; Alzet, Cupertino, CA, USA) filled with hPTH(1–34) (Bachem, Torrance, CA, USA) (11 μg [kg body weight (BW)]^–1 day^–1) dissolved in 0.9% NaCl (pH 1.2) were s.c. inserted in the interscapular region under isoflurane anesthesia (3–5% in oxygen). Osmotic pumps were incubated overnight at 37°C in 0.9% NaCl solution before the implantation. To induce hypercalcaemia, mice were treated with dihydrotachysterol (DHT) (7.5 mg mixed with 1 kg of A04 chow; Sigma‐Aldrich, St Louis, MO, USA) for 3 days. Twenty‐four mice were treated with the ET_A/ET_B receptor antagonist macitentan [27 mg (kg BW)^–1 day^–1; Actelion, Allschwil, Switzerland] orally with food, for 3 days, and then 12 mice were continued on the diet containing macitentan only and the remaining 12 mice were given food containing DHT and macitentan, for an additional 3 days.

For enrichment of distal nephron segments by COPAS (Complex Object Parametric Analyzer and Sorter), we used mice expressing enhanced green fluorescent protein under the control of the Atp6b1 gene, which encodes the H⁺‐ATPase β1 subunit (Miller et al. 2005). Four mice were used as control, four mice were made hypercalcaemic as described above, and four additional mice were fed a high calcium diet (2% Ca diet was obtained by addition of CaCl₂ to the standard diet) for 24 h.

Mice carrying activating Casr mutation (Casr ^nuf/nuf mice) (Hough et al. 2004) were purchased from Medical Research Council Harwell (Didcot, UK).

Metabolic studies

Mice were individually housed in metabolic cages (Techniplast) with drinking water and paste food (powder food:water = 1:1, w/w) provided ad libitum. Twenty‐four hour urine collection was performed after a 3 day adaptation period. For evaluating responses to diuretics, urine excreted during 1, 4 or 6 h after an i.p. bolus injection of furosemide [10 mg (kg BW)^–1; Renaudin, Aïnhoa, France], amiloride [5 mg (kg BW)^–1; Sigma‐Aldrich] or hydrochlorothiazide [50 mg (kg BW)^–1; Sigma‐Aldrich], respectively, was collected and compared with urine excreted during the respective timescale after injection of vehicle [0.9 % NaCl, 10 μl (g BW)^–1]. All diuretics were dissolved in 0.9% NaCl and 10 μl (g BW)^–1 was injected for each mouse. Urine samples were centrifuged and stored at −20°C until analysis. At the end of experiment, mice were deeply anaesthetized with an i.p. injection of ketamin [100 mg (kg BW)^–1; Streuli Pharma AG, Uznach, Switzerland]/xylasine [20 mg (kg BW)^–1; Streuli Pharma AG] and blood was obtained by retro‐orbital bleeding. Mice were then killed by cervical dislocation for the organ collection. Harvested tissue was snap frozen in liquid nitrogen and kept at −80°C. Blood samples were centrifuged for 20 min at 2000 g and serum was separated and stored at −20°C.

Urine and blood chemistry

Urinary Na⁺ and K⁺ concentrations were determined using a flame photometer (Model 420; Sherwood Scientific Ltd, Cambridge, USA). Urinary Ca²⁺, Mg²⁺, phosphate and creatinine concentrations and total serum Ca concentration were measured using a Konelab analyser (Thermo Fisher Scientific Inc., Waltham, MA, USA). Urine osmolality and pH were measured using an osmometer (Roebling, Berlin, Germany) and a pH lab 827 meter (Metrohm AG, Herisau, Switzerland), respectively. Blood Na⁺, K⁺ and Ca²⁺ concentrations were measured by blood analysis system (Epoc, Alere Inc., Waltham, MA, USA).

Microdissection of renal tubules

Left mouse kidney was briefly digested with liberase (Roche, Basel, Switzerland) as described previously (Morla et al. 2008). Renal tubules were isolated manually in accordance with morphological differences and were lysed in RNA extraction buffer from a RNeasy micro kit (Qiagen, Valencia, CA, USA).

Cell culture

mCCD_cl1 cells (i.e. highly differentiated immortalized mouse collecting duct cells) were cultured as described previously (Bustamante et al. 2008; Montesano et al. 2009). Cells were seeded on semipermeable polycarbonate filters (Transwell, Costar; Corning Inc. Corning, NY, USA) and grown to confluence for 4 days. After 24 h of serum starvation, cells were treated with medium containing 3 mm Ca (Ca concentration was adjusted by CaCl₂) or 1 mm Ca for 24 h.

COPAS sorting

The protocol for COPAS sorting was adapted from a previous study (Picard et al. 2014).

Microarray assay

Gene expression analysis was performed with a Agilent® SurePrint G3 Mouse GE 8 × 60 K Microarray (AMADID 28005; Agilent Technologies Inc., Santa Clara, CA, USA) and the following dual‐colour design: samples for DHT treated mice were labelled with Cy5, whereas control samples were labelled in Cy3 using a two‐colour labelling kit (Low Input Quick Amp Labelling Kit 5190‐2306; Agilent Technologies Inc.) adapted for small amount of total RNA (100 ng of total RNA per reaction). Hybridization was then performed on the microarray using 825 ng of each linearly amplified cRNA labelled Cy3 or Cy5 sample in accordance with the manufacturer's instructions (SureHyb Chamber; Agilent Technologies Inc.; 1650 ng of labelled extract; duration of hybridization of 17 h; 40 μL per array; temperature of 65°C). After washing in acetonitrile, slides were scanned using a G2565 C DNA microarray scanner (Agilent Technologies Inc.) with default parameters (100° PMT, 3 μm resolution) at 20°C in a free ozone concentration environment. Microarray images were analysed using Feature Extraction, version 10.7.3.1 (Agilent Technologies Inc.). Default settings were used. Four replicates per condition were used for the microarray assay.

Microarray data processing and analysis

Raw data files from Feature Extraction were imported into R with LIMMA (Smyth, 2004), an R package from the Bioconductor project (https://www.bioconductor.org) and processed: gMedianSignal and rMedianSignal data were imported, controls probes were systematically removed, and flagged probes (gIsSaturated, gIsFeatpopnOL, gIsFeatNonUnifOL, rIsSaturated, rIsFeatpopnOL, rIsFeatNonUnifOL) were set to NA. Intra‐array normalization was performed by a loess normalization, followed by a quantile normalization of both Cy3 and Cy5 channels. Next, inter‐array normalization was performed by quantile normalization on M values. A single value was obtained for each transcript by taking the mean of the summarized data for each replicated probe. Missing values were inferred using th KNN algorithm from the package ‘impute’ from R Bioconductor. Normalized data were then analysed. To assess differentially expressed genes between two groups, we started by fitting a linear model to the data using LIMMA. We then used an empirical Bayes method to moderate the SEs of the estimated log‐fold changes. The top‐ranked genes were selected with the criteria: absolute fold‐change > 2 and adjusted P value (false discovery rate) < 0.05.

Gene expression analysis

RNA extraction from whole kidneys, microdissected renal tubules, and mCCD cells was performed with TRIzol (Invitrogen, Carldbad, CA, USA), a Qiagen RNeasy micro kit and an EZNA total RNA kit (Quanta Bio, Beverly, MA, USA) in accordance with the respective manufacturer's instructions. All RNA samples were treated with DNase I (Qiagen) in accordance with the manufacturer's instructions. Reverse transcriptase PCR was performed with a Transcriptor First Strand cDNA Synthesis kit (Roche) or qScript cDNA Supermix (Quanta Bio). Quantitative PCR was performed using a Light Cycler 480 (Roche) with SYBR green master mix (Roche) and pairs of primers as listed in Table 1. mRNA expression was normalized by Gapdh mRNA expression for whole kidneys and microdissected renal tubules, as well as by Rplp0 mRNA for mCCD cells.

Table 1.

List of primers and sequences

Name of primer	Gene	Sequence
GAPDH Fw	Gapdh	CATGTTCCAGTATGACTCCACTC
GAPDH Rv		GGCCTCACCCCATTTGATGT
P0 Fw	Rplp0	AATCTCCAGAGGCACCATTG
P0 Rv		GTTCAGCATGTTCAGCAGTG
ET1 Fw	Edn1	CCGGGTCTTATCTCTGGC
ET1 Rv		AGGAACGCTTCTGACTCG
ET2 Fw	Edn2	TTTGGGAGCACAGAACCT
ET2 Rv		CATTTCATCCTTTATTACAAACATGAGT
ET3 Fw	Edn3	CCAGATGGTAAACTTCAGAGGAC
ET3 Rv		TTTAATAGATCCCATAAAGGAGTTACGTAG
ETAR Fw	Ednra	GGGTCATTAGCAACCCACA
ETAR Rv		CAGGTAGCCCTGGCTTT
ETBR Fw	Ednrb	GTCCCAGCGTTCGTAG
ETBR Rv		TGGTAGTATGCCTCCG
NKCC2 Fw	Slc12a1	CCAAGCGCTCCGTATTATAACCA
NKCC2 Rv		ACCAGGGGCCACAGTCACATTC
CEBPB Fw	Cebpb	CTGAGTAATCACTTAAAGATGTTCCTG
CEBPB Rv		CACTTTAATGCTCGAAACGG
CEBPD Fw	Cebpd	ATCTAGGGACAACGTGTAGATT
CEBPD Rv		AGTATCTTAGTAGAAAGGCACATAGC

Protein expression analysis

Western blot analysis for ENaC subunits was performed with total protein extracted from a half‐whole kidney as described previously (Tokonami et al. 2013). For NCC and NKCC2, a membrane protein fraction was prepared. Half‐whole kidney was homogenized in buffer containing 0.855% sucrose, 50 mm Tris‐Hepes, 1 μm EDTA, and protease and phosphatase inhibitors. Samples were centrifuged for 15 min at 3000 g at 4°C, and then the recovered supernatants were centrifuged for 30 min at 17 000 g at 4°C. Pellets were resuspended in RIPA buffer. Protein samples were mixed with the Laemmli buffer containing dithiothreitol and samples for detecting ENaC subunits were heated for 5 min at 95°C. Forty to fifty micrograms of protein was loaded for SDS‐PAGE and transferred to polyvinylidene fluoride or nitrocellulose membrane. Then, membranes were blocked with 5% skim milk or 2% bovine serum albumin (Sigma‐Aldrich) in PBS or Tris‐buffered saline and incubated with anti‐NKCC2 (dilution 1:5000), ‐NCC (dilution 1:1000), ‐pT53NCC (dilution 1:5000), ‐pT58NCC (dilution 1:5000), ‐alpha ENaC (dilution 1:3000), ‐beta ENaC (dilution 1:10 000) or ‐gamma ENaC (dilution 1:1000) rabbit antibody overnight at 4°C. Membrane was washed with PBS or Tris‐buffered saline and incubated with HRP‐conjugated rabbit IgG (dilution 1:10 000; Bio‐Rad, Hercules, CA, USA) and signals were revealed with ECL Western Blotting substrate (Thermo Fisher Scientific Inc.). Total or membrane protein expression was normalized with the signal obtained with β‐actin expression or Coomassie brilliant blue staining, respectively. Immunohistochemical analysis was performed as described previously (Tokonami et al. 2013) with rabbit anti‐beta ENaC (dilution 1:10 000) and goat anti‐AQP2 (dilution 1:200; Santa Cruz Biotechnology, Santa Cruz, CA, USA) antibodies or rabbit anti‐NKCC2 (dilution 1:2000) antibody.

Statistical analysis

Data are reported as the mean ± SE. A two‐tailed unpaired Student's t‐test was used for the statistical analysis. P < 0.05 was considered statistically significant.

Results

Vitamin D‐induced acute hypercalcaemia leads to polyuria and natriuresis in mice

Hypercalcaemia was induced by DHT in mice receiving exogenous PTH at a constant rate throughout the experiment to avoid a hypercalcaemia‐induced decrease in PTH level. Mice fed a standard laboratory diet (i.e. controls) had normal blood calcium and urinary calcium excretion levels (Fig. 1 A and B). The blood ionized calcium concentration was ∼2‐fold higher in DHT‐treated than in control mice on the third day of treatment (Fig. 1 A). Mice treated with DHT showed a significant and progressive increase in urinary calcium excretion from the first to the third day of treatment (Fig. 1 B). Hypercalcaemic mice had a higher urinary volume (Fig. 1 C) and magnesium excretion (Table 2) and a lower urinary pH and osmolality than control mice (Table 2). On the first and second day of treatment with DHT, urinary excretion of sodium was similar in DHT‐treated and control mice (Fig. 1 D). However, on the third day, DHT‐treated mice excreted significantly more sodium than controls, despite a lower BW and food intake (Fig. 1 D and Table 2). The urinary sodium to potassium ratio in DHT‐treated mice was higher than that in control mice (Table 2).

Figure 1. Effect of a 3‐day treatment with DHT on blood calcium concentration and urinary calcium, water and sodium excretions in PTH clamped micex.

A, blood ionized calcium concentration in control (n = 6) and mice treated with DHT for 3 days (n = 5). B, 24 h urinary calcium excretion in control (open circle) and DHT‐treated (closed circle) mice (n = 12). C, 24 h urine and (D) urinary sodium excretion in control and DHT‐treated mice (n = 12). Values are expressed as the mean ± SE. Student's t‐test, ^* P < 0.05, ^** P < 0.01, ^*** P < 0.001.

Table 2.

Twenty‐four hour urine and plasma chemistry in control and mice treated with DHT for 3 days

	Control	DHT	p
Body weight (g)	25.2 ± 0.44 (12)	22.8 ± 0.51 (12)	< 0.001
Food intake (g/24 h)	6.83 ± 0.23 (12)	4.39 ± 0.60 (12)	< 0.01
Urine
Volume (ml/24 h)	1.37 ± 0.01 (12)	2.04 ± 0.15 (12)	< 0.01
Osmolality (mosm/kg H2O)	1767 ± 92.8 (12)	1250 ± 65.5 (12)	< 0.01
pH	6.24 ± 0.11 (12)	5.24 ± 0.09 (7)	< 0.001
Na+/Creatinine (mmol/mmol)	36.0 ± 2.33 (12)	45.2 ± 3.49 (12)	< 0.05
K+/Creatinine (mmol/mmol)	80.8 ± 7.46 (12)	69.5 ± 6.85 (12)	NS
Na+/K+ (mmol/mmol)	0.48 ± 0.04 (12)	0.70 ± 0.04 (12)	< 0.001
Cl−/Creatinine (mmol/mmol)	63.6 ± 5.16 (9)	68.2 ± 10.6 (7)	NS
Ca2+/Creatinine (mmol/mmol)	0.38 ± 0.06 (11)	6.61 ± 0.57 (12)	< 0.001
Mg2+/Creatinine (mmol/mmol)	3.72 ± 0.16 (8)	8.99 ± 0.24 (8)	< 0.001
Pi/Creatinine (mmol/mmol)	21.4 ± 3.8 (11)	48.6 ± 8.0 (12)	< 0.01
Creatinine excretion rate (μmol/24 h)	5.13 ± 0.48 (12)	5.68 ± 0.53 (9)	NS
Blood/plasma
Na+ (mm)	154 ± 0.71 (6)	151 ± 1.22 (10)	NS
K+ (mm)	4.55 ± 0.24 (6)	4.59 ± 0.15 (10)	NS
Ca2+ (mm)	1.31 ± 0.03 (6)	2.22 ± 0.09 (10)	< 0.001
Total Ca (mm)	2.34 ± 0.04 (12)	4.51 ± 0.17 (12)	< 0.001
Mg2+ (mm)	1.27 ± 0.04 (5)	1.51 ± 0.06 (5)	< 0.01
Creatinine (μm)	11.4 ± 0.86 (5)	12.5 ± 0.73 (5)	NS

Values are the mean ± SEM. P value was obtained by Student's t‐test. NS, not significant.

Changes in expression of ET isoforms and receptors along the nephron during vitamin D‐induced hypercalcaemia

We analysed the relative mRNA levels of Edn1, Edn2, Edn3, Ednra and Ednrb in control and DHT‐treated mice. In control mice, Edn1 was highly expressed in medullary and cortical TAL, DCT, connecting tubule, and cortical and outer medullary collecting ducts (CCD and OMCD, respectively) (Fig. 2 A). Edn2 mRNA was not detected (data not shown). Edn3 expression was limited to the more distal nephron segments: DCT, CNT, CCD and OMCD (Fig. 2 B). Ednra expression was high in medullary and cortical TAL, DCT and CNT (Fig. 2 C). Ednrb expression pattern was similar to that of Edn1 (Fig. 2 D). Compared to controls, Edn1 expression was significantly higher in DCT, CNT and CCD of DHT‐treated mice (Fig. 2 A). No difference in the expression of Edn3, Ednra and Ednrb was detected between control and DHT‐treated mice (Fig. 2 B, C and D).

Figure 2. Relative mRNA level of ET‐1 and ‐3 and ET receptors in control and DHT‐treated mice.

Relative mRNA level of (A) ET‐1 (Edn1), (B) ET‐3 (Edn3), (C) ET_A receptor (Ednra) and (D) ET_B receptor (Ednrb) in microdissected proximal convoluted tubule (PCT), proximal straight tubule (PST), medullary thick ascending limb (MTAL), cortical thick ascending limb (CTAL), DCT, CNT, CCD, OMCD from control (white bar) and mice treated with DHT for 3 days (black bar). mRNA level of each gene was normalized with Gapdh mRNA. Values are mean±SE. n = 6 mice, Student's t‐test, ^** P < 0.01, ^*** P < 0.001.

Identification of the mechanism responsible for increased ET‐1 expression during vitamin D‐induced hypercalcaemia

To identify the genes involved in the increased ET‐1 expression during vitamin D‐induced hypercalcaemia, we performed a microarray analysis on CNT/CCD isolated from DHT‐treated and control mice. One hundred and fifty‐seven transcripts, including Cyp24a1 and Gdf15, were more than 2‐fold more abundant in the CNT/CCD from DHT‐treated mice than in those from control mice (see Supporting information, Table S1). We found that Cebpb and Cebpd expression was strongly upregulated in CNT/CCD from DHT‐treated mice. These genes encode CCAAT/enhancer‐binding proteins β and δ, respectively, which are responsible for the increased ET‐1 expression induced by high glucose in human endothelial cells (Manea et al. 2013). This result was further confirmed by a quantitative PCR on dissected tubular segments. Specifically, we found that, compared to controls, DHT‐treated mice exhibited significantly higher Cebpb mRNA levels in DCT, CNT, CCD and OMCD, as well as higher Cebpd expression in DCT (Fig. 3 A and B).

Figure 3. Relative gene expression of ET‐1 and CEBP β and δ in mCCD cells and mouse distal nephron.

Relative mRNA expression of (A) CEBP β (Cebpb) and (B) CEBP δ (Cebpd) in microdissected DCT, CNT, CCD and OMCD from control mice (white bar) and mice treated with DHT for 3 days (black bar) (n = 6 mice). mRNA expression of each gene was normalized to Gapdh mRNA expression. Relative mRNA expression of ET‐1 (Edn1) in (C) distal tubules enriched from the kidney of control (white bar), mice fed a high Ca diet (grey bar) or treated with DHT (black bar) (n = 3 or 4 mice), (D) control mCCD cells (white bar) and cells treated with 3 mm Ca (grey bar) for 24 h (n = 4 independent experiments) and (E) the whole kidney harvested from wild‐type and CaSR^nuf/nuf mice (n = 3 or 4 mice). mRNA expression of each gene was normalized to Gapdh or Rplp0 mRNA expression. Values are the mean ± SE. Student's t‐test, ^* P < 0.05.

To determine whether the increased ET‐1 expression was a result of the effect of DHT or high extracellular calcium, we examined the effect of hypercalcaemia induced either by a high Ca diet or by DHT. High Ca diet induced a significant increase of blood ionized calcium level (control: 1.31 ± 0.03 mmol/L, high Ca diet: 1.67 ± 0.09 mmol/L, P = 0.003, n = 4 or 5 mice). As expected, DHT‐treated mice increased Edn1 expression in the distal nephron; however, Edn1 expression was significantly decreased in the distal nephron of mice fed a high Ca diet (Fig. 3 C). To assess whether this effect could be reproduced in an in vitro model, the effect of high extracellular calcium on immortalized mouse CCD (mCCD) cells was tested. Edn1 expression was significantly downregulated by 3 mm extracellular Ca (Fig. 3 D). Casr ^nuf/nuf mice, where Casr is constitutively activated, showed Edn1 expression in the kidney identical to that in wild‐type littermate mice (Fig. 3 E).

Effect of ET system inhibition during vitamin D‐induced hypercalcaemia

To investigate the role of ET‐1 during vitamin D‐induced hypercalcaemia, we examined the effect of macitentan, an ET_A/ET_B receptor antagonist. Mice treated with macitentan exhibited identical urinary calcium excretion and blood ionized calcium concentration as controls (Fig. 4 A and B). These values were also similar between macitentan+DHT‐treated mice and those given DHT only (Fig. 4 A and B). This suggests that inhibition of ET_A/ET_B receptors had no significant effect on calcium homeostasis or on the ability of DHT to induce hypercalcaemia in mice. Macitentan‐treated mice showed an increased urinary volume compared to controls (Fig. 4 C). However, urinary volume was even higher in macitentan+DHT‐treated mice (Fig. 4 C). The latter had also lower urinary osmolality than mice treated with macitentan only (Table 3), suggesting that polyuria and urine concentration defects during hypercalcaemia were independent of the ET system. Sodium excretion was identical in macitentan‐treated and macitentan+DHT‐treated mice (Fig. 4 D). Furthermore, BW, food intake and the urinary sodium to potassium ratio were similar between macitentan+DHT‐treated mice and those given macitentan alone (Table 3).

Figure 4. Effect of the ET_A/ET_B receptor inhibitor macitentan on renal sodium and water excretions in control and DHT‐treated mice.

Control and DHT‐treated mice, and macitentan‐ and macitentan+DHT‐treated mice were studied on two separate experiments. A, 24 h urinary calcium excretion in control mice (open circle) and mice treated with DHT (closed circle), macitentan (open square) and macitentan+DHT (closed square) (n = 12). B, blood ionized calcium concentration in control mice and mice treated with DHT, macitentan or macitentan+DHT (n = 5 or 6). C, 24 h urine and (D) urinary sodium excretion in control mice, and mice treated with DHT, macitentan or macitentan+DHT (n = 12). E, relative ET‐1 mRNA (Edn1) in whole kidney from control mice and mice treated with DHT, macitentan or macitentan+DHT (n = 5–6). Values are expressed as the mean ± SE. Student's t‐test, ^* P < 0.05, ^** P < 0.01, ^*** P < 0.001. NS, not significant.

Table 3.

Twenty‐four hour urine and plasma chemistry in Macitentan‐ and Macitentan+DHT‐treated mice

	Macitentan‐only	Macitentan+DHT	P
Body weight (g)	24.1 ± 0.37 (12)	23.52 ± 0.21 (12)	NS
Food intake (g/24 h)	7.71 ± 0.21 (12)	7.19 ± 0.25 (12)	NS
Urine
Volume (ml/24 h)	1.94 ± 0.14 (12)	2.61 ± 0.22 (12)	< 0.05
Osmolality (mosm kg−1 H2O)	1853 ± 116 (12)	1503 ± 79.8 (12)	< 0.05
pH	7.21 ± 0.07 (11)	5.78 ± 0.07 (9)	< 0.001
Na+/Creatinine (mmol/mmol)	42.7 ± 3.3 (12)	42.8 ± 3.7 (12)	NS
K+/Creatinine (mmol/mmol)	66.5 ± 2.6 (12)	68.6 ± 3.4 (12)	NS
Na+/K+ (mmol/mmol)	0.59 ± 0.01 (12)	0.60 ± 0.02 (12)	NS
Ca2+/Creatinine (mmol/mmol)	0.42 ± 0.07 (12)	6.43 ± 0.23 (12)	<0.001
Mg2+/Creatinine (mmol/mmol)	2.69 ± 0.19 (5)	8.33 ± 0.25 (5)	<0.001
Creatinine excretion rate (μmol/24 h)	6.55 ± 0.33 (12)	7.03 ± 0.38 (12)	NS
Blood/plasma
Na+ (mm)	153 ± 0.58 (5)	153 ± 1.86 (6)	NS
K+ (mm)	4.24 ± 0.08 (5)	4.45 ± 0.22 (6)	NS
Ca2+ (mm)	1.25 ± 0.01 (5)	2.28 ± 0.22 (6)	<0.001
Total Ca (mm)	2.40 ± 0.13 (8)	4.29 ± 0.22 (9)	<0.001
Mg2+ (mm)	1.26 ± 0.10 (5)	1.57 ± 0.06 (5)	<0.05
Creatinine (μm)	12.4 ± 0.61 (5)	11.3 ± 1.05 (5)	NS

Values are the mean ± SEM. P value was obtained by Student's t‐test. NS, not significant.

We performed Edn1 gene expression analysis on whole kidneys harvested from mice treated with macitentan only or macitentan+DHT. A comparable increase in Edn1 mRNA levels was observed between DHT‐ and macitentan+DHT‐treated mice, as well as controls and macitentan‐treated mice, respectively (Fig. 4 E). No significant difference in Ednra and Ednrb mRNA levels was observed between the groups (data not shown).

Identification of distal sodium transporter(s) affected by vitamin D‐induced hypercalcaemia

Having observed increased Edn1 expression in DCT, CNT, and CCD in hypercalcaemic mice, we investigated whether ET‐1 could affect one or more sodium transporters expressed in one or several segments of the distal nephron.

ENaC

To assess the effect of vitamin D/hypercalcaemia‐elicited ET‐1 expression on ENaC activity in vivo, we studied the acute response to amiloride [5 mg (kg BW⁻¹)]. As expected, DHT‐treated mice excreted significantly more sodium in the urine than control mice during the 4 h after injection of normal saline solution (Fig. 5 A). The opposite was observed after amiloride administration: urinary sodium excretion was higher in control than in DHT‐treated mice (Fig. 5 A). This finding suggested a significant reduction of ENaC activity in vitamin D‐treated hypercalcaemic mice. To examine whether ENaC was controlled by ET‐1 during vitamin D‐induced hypercalcaemia, we performed the same amiloride test on macitentan‐treated mice. Sodium excretion was similar in mice treated with macitentan+DHT or macitentan alone (Fig. 5 A) and was consistent with normal ENaC activity under this condition.

Figure 5. Amiloride‐sensitive urinary sodium excretion and expression profile of ENaC subunits during vitamin D‐induced hypercalcaemia.

Urinary sodium (A) and calcium and magnesium excretions (B) 4 h after a bolus injection of vehicle (0.9% NaCl) or amiloride [5 mg (kg BW⁻¹)] in control (white bar), DHT‐treated (black bar), macitentan‐treated (light grey bar) and macitentan+DHT‐treated (dark grey bar) mice (n = 5–7). Total protein expression of (C) alpha, (D) beta and (E) gamma ENaC subunits and relative protein expression of (F) alpha, (G) beta and (H) gamma ENaC subunits obtained by densitonometry quantification in control and mice treated with DHT, macitentan and macitentan+DHT. Actin was used for loading control. Values are expressed as the mean ± SE. Student's t‐test, ^* P < 0.05, ^** P < 0.01, ^*** P < 0.001. NS, not significant.

Amiloride decreased urinary calcium and magnesium excretion in all mice (Fig. 5 B). Specifically, DHT‐ and macitentan+DHT‐treated mice showed a similar decrease in urinary calcium and magnesium levels after amiloride injection.

Western blot analysis revealed that the expression of uncleaved α, γENaC and cleaved αENaC subunits were identical in DHT‐treated and control mice, as well as in macitentan+DHT‐ and macitentan‐treated mice (Fig. 5 C, E, F and H). βENaC subunit protein expression level was higher in DHT‐treated and macitentan+DHT‐treated mice than control and macitentan‐treated mice, respectively (Fig. 5 D and G). However, subcellular localization of the βENaC subunit did not change between groups both in DCT and CNT/CCD (Fig. 6). In summary, our data indicate that ENaC‐mediated sodium transport is functionally inhibited in an ET‐1‐dependent manner in vitamin D‐treated hypercalcaemic mice.

Figure 6. Subcellular localization of β ENaC subunit (red) in control, DHT‐, macitentan‐ and macitentan+DHT‐treated mouse kidneys.

Connecting tubles and collecting ducts are identified by immunolabelling for AQP2 (green). Distal convoluted tubules are identified in tubules positive for βENaC and negative for AQP2 staining (tubules marked by asterisks). Images were acquired and analysed by confocal microscopy. Nucleus is stained with 4′,6‐diamidino‐2‐phenylindole (blue). Scale bar = 50 μm.

NCC and pendrin/sodium‐dependent chloride bicarbonate exchanger (NDCBE)

Sodium reabsorption is also mediated by NCC and pendrin/NDCBE expressed at the apical membrane of DCT and beta‐intercalated cells in CNT/CCD, respectively. To assess sodium handling by these transporters, we examined the diuretic response to a bolus injection of hydrochlorothiazide [HCTZ; 50 mg (kg BW⁻¹)], an inhibitor of NCC and NDCBE, in control and DHT‐treated mice. DHT‐treated mice excreted more sodium than control mice during the 6 h after HCTZ injection (Fig. 7 A). This indicated an increase in HCTZ‐sensitive sodium transport during vitamin D‐induced hypercalcaemia. HCTZ decreased urinary calcium and magnesium excretion in all mice (Fig. 7 B). The level of protein expression of total NCC, phosphorylated NCC at position T53 and T58 was significantly higher in DHT‐treated than in control mice (Fig. 7 C and E), in agreement with the functional test. Moreover, phosphorylated NCC, on both T53 and T58 residues, was more abundant in DHT‐treated mice than in control mice (Fig. 7 C and E). The same trend was observed in mice treated with macitentan+DHT vs. macitentan alone (Fig. 7 D and F). No significant differences in the mRNA expression of pendrin and NDCBE were observed between control and DHT‐treated mice (data not shown).

Figure 7. HCTZ‐sensitive sodium excretion and NCC expression during vitamin D‐induced hypercalcaemia.

Urinary sodium (A) and calcium and magnesium excretions (B) 6 h after a bolus injection of vehicle (0.9% NaCl) or HCTZ [50 mg (kg BW⁻¹)] in control (white bar) and DHT‐treated (black bar) mice (n = 4–6). C, membrane protein expression of total NCC (tNCC), and phosphorylated NCC at residues T53 (pT53NCC) and T58 (pT58NCC) (C) in control and DHT‐treated mice and (D) in macitentan and macitentan+DHT‐treated mice. Relative protein expression of total and phosphorylated NCC obtained by densitometry quantification (E) in control and DHT‐treated mice and (F) in macitentan and macitentan+DHT‐treated mice. Commassie brilliant blue staining was used for loading control. Values are expressed as the mean ± SE. Student's t‐test, ^* P < 0.05, ^** P < 0.01.

NKCC2

Hypercalcaemic rats exhibit reduced NKCC2 expression, which has been suggested to cause natriuresis during hypercalcaemia (Wang et al. 2002 a). To test whether this was the case in our model, we examined the response to a bolus injection of furosemide [10 mg (kg BW⁻¹)], a NKCC2 inhibitor, in control and DHT‐treated mice. As shown in Fig. 8 A, sodium excretion during the 1 h after furosemide injection was similar in all mice, indicating that NKCC2 activity in control and DHT‐treated mice was identical. No significant differences in mRNA, membrane protein level or subcellular localization of NKCC2 were observed between control and DHT‐treated mice (Fig. 8 B and E).

Figure 8. Furosemide‐sensitive sodium excretion and NKCC2 expression during vitamin D‐induced hypercalcaemia.

A, urinary sodium excretion 1 h after a bolus injection of furosemide [10 mg (kg BW⁻¹)] in control (white bar) and DHT‐treated (dark bar) mice (n = 6). B, relative mRNA expression of NKCC2 (Slc12a1) in medullary thick ascending limb (MTAL) and cortical thick ascending limb (CTAL) from control (white bar) and DHT‐treated (black bar) mice (n = 6). C, membrane protein expression of NKCC2 and (D) relative protein expression obtained by densitometry quantification in control and DHT‐treated mice. Commassie brilliant blue staining was used for loading control. E, subcellular localization of NKCC2 (red) in control and DHT‐treated mouse kidney. Images were acquired and analysed by confocal microscopy. The nucleus is stained with 4′,6‐diamidino‐2‐phenylindole (blue). Scale bar = 50 μm. Values are expressed as the mean ± SE. Student's t‐test, NS, not significant. [Color figure can be viewed at wileyonlinelibrary.com]

Discussion

The present study aimed to identify the mechanism underlying the effect of acute hypercalcaemia on ion and water transport in the kidney. For this purpose, we first developed a hypercalcaemic mouse model in which PTH‐supplemented mice received oral treatment with DHT. As expected, DHT induced hypercalcaemia in mice after 3 days, probably by increasing intestinal calcium absorption and renal calcium reabsorption. Vitamin D‐treated hypercalcaemic mice exhibited increased urinary volume, urine concentration defect and renal sodium loss. These outcomes were consistent with previous studies in humans and other animal models (Wolf & Ball, 1949; Levitt et al. 1957; DiBona, 1971; Fuleihan et al. 1998; Adami & Parfitt, 2000; Battula et al. 2012).

Vitamin D‐induced hypercalcaemic mice exhibited increased ET‐1 mRNA expression in the kidney, in line with observations made by Shiraishi et al. (2003) in rats. So far, the mechanism responsible for overexpression of ET‐1 during vitamin D‐induced hypercalcaemia was unresolved and, especially, the effect of vitamin D on ET‐1 expression had not been investigated. The present study provides evidence indicating that the elevated extracellular calcium concentration may not be a principal cause of increased ET‐1 during vitamin D‐induced hypercalcaemia. Although Shiraishi et al. (2003) showed that activation of CaSR could increase ET‐1 promoter activity in transfected COS7 cells, our in vivo and in vitro data indicate that a high extracellular calcium concentration actually decreases ET‐1 mRNA expression both in the distal nephron and in mCCD cells that constitutively express Casr. These contrasting results may be a result of differences between the transient overexpression system and in vivo conditions or a differentiated stable cell line system generated from immortalized cells. Similarly, the effect of CaSR activation may vary depending on cell type; for example, COS7 cell is a fibrobrast‐like non‐epithelial cell line derived from the monkey kidney, whereas mCCD cells are epithelial cells (Jsensen et al. 1964). CaSR is expressed in various renal cell types in the distal nephron (Graca et al. 2016). In the present study, the vitamin D/hypercalcaemia‐induced increase in ET‐1 expression was restricted to DCT, CNT and CCD, and none occurred in TAL, despite the much higher expression of CaSR in medullary thick ascending limb and cortical thick ascending limb than in any other tubular segment (Loupy et al. 2012; Graca et al. 2016). Moreover, ET‐1 mRNA expression did not significantly differ between mice expressing an activating mutation of Casr and wild‐type mice. Overall, our findings suggest that Casr is probably not involved in the control of ET‐1 expression. The mechanism underlying the reduced ET‐1 expression by hypercalcaemia on a high Ca diet remains to be investigated.

We found increased Cebpb and Cebpd mRNA levels in the distal nephron of DHT‐treated hypercalcaemic mice. This is in line with previous studies showing that 1,25(OH)₂‐vitamin D positively regulated CCAAT/enhancer‐binding protein (CEBP) β gene and protein expression in COS7 cells (Dhawan et al. 2005) and that the mRNA and protein expression of CEBP β and δ subunits could be upregulated in primary keratinocytes in response to a high extracellular calcium level (Oh & Smart, 1998). Therefore, we hypothesize that vitamin D controls ET‐1 expression via activation of the CEBP pathway. The detailed mechanism underlying the regulation of vitamin D/hypercalcaemia‐induced ET‐1 via this pathway remains to be determined.

ET‐1 mRNA overexpression was limited to DCT, CNT and CCD in vitamin D‐treated hypercalcaemic mice. Therefore, we hypothesized that, during vitamin D‐induced acute hypercalcaemia, ET‐1 exerted its effect primarily via autocrine signalling on DCT, CNT and/or CCD, or via paracrine signalling on neighbouring cells.

We showed that inhibition of ET_A/ET_B receptors by macitentan suppressed vitamin D/hypercalcaemia‐elicited natriuresis without affecting renal calcium/magnesium handling or the expression of ET‐1 and its receptors. This suggests that activation of the ET system can directly inhibit renal sodium transport during vitamin D‐induced hypercalcaemia. By contrast, we observed that macitentan did not affect vitamin D/hypercalcaemia‐induced polyuria or defects in urine concentration, suggesting that these disorders were independent of ET‐1. Our finding was in line with a previous study showing that the activation of ET_A or ET_B receptors in vivo did not induce polyuria in male rats (Nakano & Pollock, 2009). The mechanism involved in macitentan‐induced polyuria remains unknown. A single study previously demonstrated that genetic ablation of ET_A receptor in mice results in the reduction of AVP‐induced cAMP accumulation in the CD (Ge et al. 2005). Therefore, it is possible that macitentan altered AVP signalling via inactivation of ET_A receptor.

Vitamin D‐induced hypercalcaemic mice have reduced amiloride‐sensitive sodium transport, indicating that natriuresis could result from lower sodium reabsorption through ENaC. Furthermore, amiloride‐sensitive sodium transport was normalized after treatment by macitentan. This is in agreement with previous studies using genetically modified animals or pharmacological tools to show the inhibitory role of ET‐1 on ENaC activity (Bugaj et al., 2008, 2012). Moreover, given the lack of difference in ENaC expression between control and DHT‐treated mice, the activated ET system probably controls ENaC activity at a post‐transcriptional level. Indeed, Bugaj et al. (2012) demonstrated that ET‐1 lowered open probability of ENaC within 5 min in isolated split‐open rat CCD, indicating that ET‐1 may control ENaC without affecting its expression level (Bugaj et al. 2008). Another study suggested that the reduction of ENaC activity induced by ET‐1 could be mediated by the activation of betaPix/14‐3‐3/Nedd4‐2 pathway (Pavlov et al. 2010). Whether vitamin D/hypercalcaemia‐induced ET‐1 controls ENaC activity via the same pathway remains to be determined.

We observed that vitamin D‐induced hypercalcaemia increased HCTZ‐sensitive sodium transport and NCC expression. This effect appeared to be independent of the ET system because inhibition of ET receptors by macitentan did not affect NCC expression and phosphorylation. The mechanism underlying vitamin D/hypercalcaemia‐induced increase in NCC expression is unknown. One possible explanation is that Gdf15, which has been shown to control cell proliferation in the distal nephron (Van Huyen et al. 2008) and is highly increased in the distal nephron of DHT‐treated mice, might be involved in this effect. Because DCT may become hyperplastic in response to environmental stimuli or homeostatic challenges (Ellison et al. 1989; Loffing et al. 1995; Wagner et al. 2008), we speculate that the increase in NCC expression could be a result of the increased DCT mass in DHT‐treated mice. Further studies are required to investigate the possible role of vitamin D on DCT remodelling.

Finally, in contrast to previous studies conducted in rats, we did not observe significant differences in furosemide‐sensitive sodium transport or NKCC2 expression between control and vitamin D‐treated hypercalcaemic mice. This discrepancy may be because previous studies examined the effect of DHT for 8 days, which is a substantially longer time than in our case. Moreover, a recent study has shown that a patient with a mutation in Umod, which encodes a protein involved in the regulation of NKCC2 in TAL, exhibited reduced NKCC2 expression at the same time as having an increased natriuretic response to furosemide. This finding indicates the inverse relationship that may exist between NKCC2 expression and its function (Labriola et al. 2015). Thus, it remains to be determined whether lower NKCC2 levels in hypercalcaemic rats cause a decreased sodium chloride reabsorption in the TAL.

In summary, vitamin D‐induced hypercalcaemia activates ET‐1 synthesis in the distal nephron, which decreases sodium reabsorption, presumably through a reduction in ENaC activity. By contrast, vitamin D/hypercalcaemia‐induced polyuria and urine concentration defects are probably not mediated by increased ET‐1. Further studies are required to reveal the mechanism underlying these defects during hypercalcaemia.

Additional information

Competing interest

The authors declare that they have no competing interests.

Author contributions

NT, EF and PH conceived and designed the experiments. NT, LC, IM and GM collected, assembled, analysed and interpreted data. JL provided antibodies. NT, EF, JL and PH drafted and/or critically revised the article. All authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work.

Funding

This work is supported by grant from Agence Nationale de la Recherche (ANR‐12‐BSV1‐0031‐01 to PH). NT is supported by funding from Swiss National Science Foundation Early and Advanced PostdocMolibity Fellowship (P2LAP3_151782 and P300P3_158521).

Supporting information

Disclaimer: Supporting information has been peer‐reviewed but not copyedited.

Table S1. Microarray analysis data. Data obtained by microarray analysis on COPAS‐sorted connecting tubules and collecting ducts from control and DHT‐treated mice.