Differential parathyroid and kidney Ca²⁺-sensing receptor activation in autosomal dominant hypocalcemia 1

Summary

Background

Parathyroid Ca²⁺-sensing receptor (CaSR) activation inhibits parathyroid hormone (PTH) release, while activation of renal CaSRs attenuates Ca²⁺ transport and increases expression of the pore-blocking claudin-14. Patients with autosomal dominant hypocalcemia 1 (ADH1), due to activating CASR mutations, exhibit hypocalcemia but not always hypercalciuria (elevated Ca²⁺ in urine). The latter promotes nephrocalcinosis and renal insufficiency. Although CaSRs throughout the body including the kidney harbor activating CASR mutations, it is not understood why only some ADH1 patients display hypercalciuria.

Methods

Activation of the CaSR was studied in mouse models and a ADH1 patient. In vitro CaSR activation was studied in HEK293 cells.

Findings

Cldn14 showed blood Ca²⁺ concentration-dependent regulation, which was absent in mice with kidney-specific Casr deletion, indicating Cldn14 is a suitable marker for chronic CaSR activation in the kidney. Mice with a gain-of-function mutation in the Casr (Nuf) were hypocalcemic with low plasma PTH levels. However, renal CaSRs were not activated at baseline but only after normalizing blood Ca²⁺ levels. Similarly, significant hypercalciuria was not observed in a ADH1 patient until blood Ca²⁺ was normalized. In vitro experiments indicate that increased CaSR expression in the parathyroid relative to the kidney could contribute to tissue-specific CaSR activation thresholds.

Interpretation

Our findings suggest that parathyroid CaSR overactivity can reduce plasma Ca²⁺ to levels insufficient to activate renal CaSRs, even when an activating mutation is present. These findings identify a conceptually new mechanism of CaSR-dependent Ca²⁺ balance regulation that aid in explaining the spectrum of hypercalciuria in ADH1 patients.

Funding

Erasmus+ 2018/E+/4458087, the Canadian Institutes for Health research, the Novo Nordisk Foundation, the Beckett Foundation, the Carlsberg Foundation and Independent Research Fund Denmark.

Research in context.

Evidence before this study

The calcium-sensing receptor (CaSR) is a key regulator of systemic calcium balance. Parathyroid CaSR activation reduces parathyroid hormone (PTH) secretion and activation of the renal CaSR reduces tubular calcium reabsorption. Patients with activating CASR mutations exhibit autosomal dominant hypocalcemia type 1 (ADH1), a disease characterized by low blood calcium concentrations and inappropriately low serum PTH concentrations, consistent with overactivation of the parathyroid CaSR. Intriguingly, despite robust CaSR expression in the kidney, increased urinary calcium excretion is sometimes, but not always observed in ADH1 patients. Importantly, calcium wasting in these patients can, if persistent, cause kidney stone formation, renal calcium deposits and kidney disease. This complicates clinical management. Although CaSRs throughout the body are activated by increased extracellular calcium concentrations, why ADH1 patients have reduced PTH but not always urinary calcium wasting is not clear, despite abundant CaSR expression in both the kidney and the parathyroid. It is generally assumed that signaling via CaSRs in ADH1 patients carrying gain-of-function CaSR mutations is increased both in the parathyroid and in the kidney.

Added value of this study

In the present study, we combine investigations in a preclinical mouse model of ADH1 with in vitro experiments and data from a patient case. In a gain-of-function CaSR mouse model and in mice treated with specific compounds to activate the CaSR, we show that renal CaSRs are not activated at baseline in contrast to what would be expected. We provide evidence from this mouse model that this occurs because CaSR overactivity in the parathyroid gland reduces PTH and circulating serum calcium levels, which is insufficient to activate the renal CaSR. Mechanistically this may occur, at least in part, due to a higher level of CaSR expression in the parathyroid relative to the kidney.

Implications of all the available evidence

Our findings suggest an intricate interplay between parathyroid and renal CaSRs and identify a conceptually new mechanism for CaSR regulated maintenance of calcium homeostasis. Furthermore, these findings are of relevance for understanding why renal calcium wasting is not consistently observed in all patients with ADH1 and may aid in their clinical management.

Alt-text: Unlabelled box

Introduction

Plasma ionized calcium (Ca²⁺) is kept within a narrow concentration range of 1.10–1.35 mM (4.4–5.4 mg/dL).¹ Ca²⁺ activates the Ca²⁺- sensing receptor (CaSR) in the parathyroid, thereby reducing the release of parathyroid hormone (PTH) which helps reduce plasma Ca²⁺ concentration.2, 3, 4 Mechanistically this occurs as PTH is a key stimulator of Ca²⁺ reabsorption from the renal tubule, resorption from bone, and a main driver of active vitamin D metabolite 1,25(OH)₂D₃ formation. 1,25(OH)₂D₃ itself is a potent stimulator of intestinal Ca²⁺ absorption and increases renal tubular Ca²⁺ reabsorption and bone resorption.¹

In addition to regulating PTH release, local CaSR activation in the kidney and intestine regulates Ca²⁺ transport through a direct response to changes in plasma Ca²⁺ levels.^1,5, 6, 7 In the kidney, the CaSR is expressed in select segments of the tubular system, with the highest expression on the basolateral membrane of the thick ascending limb of Henle's loop (TAL) and expressed at markedly lower levels at more distal sites and the early collecting system.⁶ Inhibition of the CaSR in isolated microperfused TAL tubules increases paracellular Ca²⁺ reabsorption by increasing Ca²⁺ permeability of the TAL.⁶ This suggests that the CaSR plays an important role in the regulation Ca²⁺ transport across this segment. While the direct mechanisms mediating these acute actions remains to be fully investigated, chronic CaSR activation is well known to increase expression of the pore-blocking claudin-14 (CLDN14), which is expressed at low levels under basal conditions.^8,9 In fact, CaSR activation in the kidney plays a crucial role in the defense against hypercalcemia.¹⁰ Cldn14 expression is greatly increased by chronic activation of the CaSR during chronic hypercalcemia,^8,9 while smaller changes in Cldn14 expression are observed with increased dietary Ca²⁺ content.^8,11,12 siRNA-mediated Casr knockdown attenuates Ca²⁺-dependent increases in CLDN14 expression in vitro,¹¹ suggesting that Cldn14 expression increases with chronic CaSR activation.

Gain-of-function mutations in the CASR cause autosomal dominant hypocalcemia type 1 (ADH1). ADH1 is characterized by hypocalcemia with inappropriately low plasma PTH concentrations.^13,14 Frank hypercalciuria is not always encountered at presentation but occurs commonly after treatment.13, 14, 15, 16, 17, 18 In a study with nine ADH1 patients, only one was hypercalciuric before treatment.¹⁸ In another report, two out of eleven patients were hypercalciuric at presentation.¹⁴ This suggest that the majority of ADH1 patients do not have frank hypercalciuria at presentation. Hypercalciuria increases the risk of nephrolithiasis, nephrocalcinosis and thus chronic kidney disease in patients with ADH1.¹⁹ It is not clear why some ADH1 patients present with reduced PTH levels which are indicative of CaSR activation, but not hypercalciuria. Furthermore, in a large family with ADH1, all having the same mutation, urinary Ca²⁺ excretion tended to be highest in those receiving vitamin D treatment to correct hypocalcemia.²⁰ Treatment of hypocalcemia is avoided in asymptomatic patients to prevent hypercalciuria and related sequale.¹³ As the activating mutation of the CaSR would be present in the kidney as well as the parathyroid, why these patients do not always have hypercalciuria is unclear.

We therefore studied the activation of the CaSR in a single mouse model of ADH1 (Nuf) and after administration of the calcimimetic, etelcalcetide hydrochloride.21, 22, 23 We first validated renal Cldn14 expression as a marker of renal CaSR activation by showing that Cldn14 abundance time- and concentration-dependently increased with chronic changes in serum Ca²⁺. This regulation was absent in hypercalcemic animals with targeted ablation of the Casr in the kidney. Surprisingly, we found that renal Cldn14 expression was unchanged in Nuf mice, despite an activating CaSR mutation. However, when plasma Ca²⁺ levels were increased in Nuf mice, but still lower than wild-type animals, increased renal Cldn14 expression was observed, indicating that chronic activation of the renal CaSR is not present until blood Ca²⁺ is elevated to sufficient levels to activate the receptor. Overall, our findings from this model delineate tissue-specific CaSR thresholds, where CaSR activity in the parathyroid reduces PTH secretion and hence circulating Ca²⁺ levels, which then are insufficient to activate renal CaSRs in the chronic setting, even when containing an activating mutation. These findings suggest a new mechanism of regulated Ca²⁺ balance by the interplay between parathyroid and renal CaSRs and may aid in the management of ADH1 patients and hypercalciuria-related sequelae.

Methods

Animal experiments

Nuf mice (C3;102-Casr^Nuf/H; RRID:MGI:3603347), which harbor an activating mutation (L723Q) in the fourth transmembrane region of the CaSR, have been described previously and were obtained through MRC Harwell via the European Mouse Mutant Archive.²³ Floxed Casr^tm1Mrpk/J mice¹⁰ (RRID:IMSR_JAX:030647) and the Ksp-Cre driver line (B6.Cg-Tg(Cdh16-cre)91Igr/J) (RRID:IMSR_JAX:012237)²⁴ were both obtained via JAX and crossed to generate animals with CaSR deletion specifically in kidney.

Experimental protocol 1

Experiments were performed on 30 male WT C57BL/6J mice (Janvier Labs, Le Genest-Saint-Isle, France; 12 weeks old). Animals were divided randomly into groups receiving the control diet or a diet enriched with dihydrotachysterol (DHT; dihydrotachysterol (Sigma-Aldrich, St. Louis, MO, USA)), a synthetic vitamin D analog that is activated in the liver independently of renal hydroxylation.²⁵ The following dosing strategies were applied (n = 5 per group): 0.4% w/v DHT (per 100 g of standard rodent food) for 3 days, 0.08% w/v DHT for 3 days, 0.04% w/v DHT for 3 days, 0.4% w/v DHT for 2 days or 0.4% w/v DHT for 1 day or vehicle controls. The standard rodent maintenance diet contained 0.2% Mg²⁺, 0.7% Ca²⁺ and 600 IU vitamin D3 (TD.00374; Altromin Spezialfutter GmbH & Co. KG, Lage, Germany); animals had ad libitum access to food and drinking water. Animals were subsequently anesthetized with 1.5% isoflurane (Nicholas Piramal Ltd., London, UK) and blood was drawn from the vena cava. Mice were then euthanized by cervical dislocation and kidneys were isolated as previously.²⁶ Data are shown in Figure 1.

Figure 1.

Renal Cldn14 expression is dose-dependently regulated by blood Ca²⁺ levels via the CaSR. (a) With the exception of dietary DHT administration of 0.4% w/v DHT for 1 day, all DHT treatment groups exhibited increased blood Ca²⁺ levels compared to control mice in a dose-dependent manner (n = 4–5). P < 0.0001 for vehicle vs DHT, vehicle vs DHT – 2 days, DHT vs 1:5 DHT, DHT vs 1:10 DHT, DHT vs DHT – 2 days and DHT vs DHT – 1 day; P = 0.0004 for vehicle vs 1:5 DHT; P = 0.0301 for vehicle vs 1:10 DHT; P = 0.0050 for 1:10 DHT vs DHT – 2 days; P = 0.0014 for DHT – 2 days vs DHT – 1 day. (b) Renal Cldn14 expression was significantly increased in animals treated with 0.4% w/v DHT for 3 days compared to control animals (n = 5). P < 0.0001 for vehicle vs DHT, DHT vs 1:5 DHT, DHT vs 1:10 DHT and DHT vs DHT – 1 day; P = 0.0002 for DHT vs DHT – 2 days. (c) Western blotting for CLDN14 showed that CLDN14 protein expression in the kidney was highly abundant in animals treated with 0.4% w/v DHT for 3 days and to a lesser extent in the other treatment groups. Indicated by arrow. Note an unspecific band is also present above. β-actin confirmed equal loading (n = 4). (d) Immunohistochemical staining of kidney sections for CLDN14 as indicated by arrows. 3-day 0.4% w/v DHT-treatment increased CLDN14 expression in TAL compared to control animals. Scale bar indicates 200 µm (e) Correlation between blood Ca²⁺ levels and ∆Ct values for Cldn14 (calculated as Cldn14 Ct values – 18S Ct values) using a power trend line showed lower ∆Ct values for Cldn14, and thus higher Cldn14 expression, at higher blood Ca²⁺ levels. R²=0.73. (f) After a 3-day 0.4% w/v DHT containing diet, Cldn14 expression was significantly lower in kidney-specific Casr^−/− (KS-Casr^−/−) mice compared to WT mice (n = 4–7). P = 0.0326. (g) CLDN14 and CaSR protein expression was detectable by western blotting in DHT-treated WT mice, but not in DHT-treated KS-Casr^−/− mice. β-actin confirmed equal loading (n = 4). (h) CLDN14 expression increased in the TAL of WT mice treated with 0.4% w/v DHT for 3 days, but not in TALs of KS-Casr^−/− mice. Scale bar indicates 100 µm. DHT, 0.4% w/v DHT for 3 days; 1:5 DHT, 0.08% w/v DHT for 3 days; 1:10 DHT, 0.04% w/v DHT for 3 days. Data are shown as mean ± SEM. *P < 0.05 by one-way ANOVA with Tukey's post hoc test.

Experimental protocol 2

Sex and age-matched wild type (WT) (n = 4; 3 males and 1 female) and kidney-specific Casr^−/− (n = 7; 5 males and 2 females) mice from the same litter (aged 12 weeks) were placed in individual metabolic cages in a temperature-controlled (25 °C) environment with ad libitum access to food and drinking water. After acclimation, all animals were switched to 0.4% w/v dietary DHT administration for 3 days as described above. Mice were subsequently anesthetized and euthanized as described in protocol 1. Data are shown in Figure 1.

Experimental protocol 3

Experiments were performed on sex-matched and age-matched (average age: 10 weeks) WT (n = 9 animals, 4 males and 5 females), heterozygous (Nuf/+) (n = 15 animals, 7 males and 8 females) and homozygous (Nuf/Nuf) (n = 10 animals, 7 males and 3 females). All animals were transferred to individual metabolic cages with ad libitum access to the standard rodent maintenance diet described above and drinking water. After acclimatization for 3 days, 24 h urine was collected. Animals were subsequently anesthetized and euthanized as described in protocol 1. Data are shown in Figure 2.

Figure 2.

The renal CaSR is not activated under baseline conditions in Nuf mice. (a) Calcium levels in blood were significantly lower in Nuf/+ and Nuf/Nuf mice compared to WT animals. Ionized blood Ca²⁺ was also significantly lower in Nuf/Nuf mice compared to Nuf/+ mice (n = 8–11). P <0.0001 for WT vs. Nuf/+, WT vs. Nuf/Nuf and Nuf/+ vs. Nuf/Nuf. (b) Plasma intact PTH was significantly lower in Nuf/+ and Nuf/Nuf mice than in WT mice (n = 9–12). P < 0.0001 for WT vs. Nuf/+ and WT vs. Nuf/Nuf. (c) Urinary Ca²⁺ excretion was significantly lower in Nuf/+ mice and Nuf/Nuf mice compared to WT animals, but no significant differences were detected between Nuf/+ and Nuf/Nuf mice (n = 9–13). P = 0.0219 for WT vs. Nuf/+ and P = 0.0181 for WT vs. Nuf/Nuf. (d) Fractional urinary excretion of Ca²⁺ (FE_Ca²⁺) was similar in all genotypes (n = 4–8). (e) The renal expression of Cldn14, a marker of CaSR activation in the kidney, was not significantly different between genotypes (n = 6–7). (f) The renal expression of Cyp24a1, involved in the inactivation of 1,25(OH)₂D₃, was also unaltered in Nuf mice (n = 6–9). (g) The renal expression of Cyp27b1, involved in the activation of 1,25(OH)₂D₃, was not statistically different between genotypes (n = 6–7). (h) Immunohistochemical staining of CLDN14 in the TAL was not observed in any of the genotypes. Scale bar indicates 50 µm. Data are shown as mean ± SEM. *P < 0.05 by one-way ANOVA with Tukey's post hoc test.

Experimental protocol 4

8 WT (4 males, 4 females) and 8 Nuf/+ (4 males and 4 female) mice (age 16–17 weeks) were placed on a 2%-supplemented Ca²⁺ diet (Envigo, Indianapolis, IN, USA) for 7 days. First, they received the diet in their standard cages for 3 days and were subsequently transferred to individual metabolic cages, where they continued receiving the high Ca²⁺ diet. All animals had ad libitum access to food and water and were housed in a controlled environment, as described above. After 3 days of acclimatization, 24 h urine was collected. Mice were subsequently anesthetized and euthanized under 1.5% isoflurane as described in protocol 1. Data are shown in Figure S2.

Experimental protocol 5

8 WT (4 males, 4 females) and 7 Nuf/+ (3 males, 4 females) mice (age 16–17 weeks) were housed in individual metabolic cages as described above. After acclimatization for 3 days, baseline 24 h urine was collected. All mice were then switched to a diet supplemented with 0.4% w/v DHT as describe before for 3 days; 24 h urine was collected for all days.²⁵ Mice were anesthetized and euthanized as described above. Data are shown in Figure 3.

Figure 3.

Effect of dietary DHT supplementation on Ca²⁺ balance. (a) No significant differences in blood Ca²⁺ between WT and Nuf/+ mice were present after 3 days of dietary DHT supplementation (n = 7). (b) Plasma PTH concentrations were significantly lower in Nuf/+ mice (n = 5–6). P = 0.0016. (c) Although DHT increased urinary Ca²⁺ excretion in both genotypes, this increase was delayed significantly, by one day, in Nuf/+ mice compared to WT mice (n = 7–8). P = 0.0341. (d) Fractional urinary excretion of Ca²⁺ (FE_Ca²⁺) 3 days after DHT treatment was not significantly different between genotypes (n = 7). (e) No difference in renal Cldn14 expression was present between Nuf/+ mice and WT mice after 3 days of DHT supplementation (n = 7–8). (f,g) The renal expression of Cyp24a1 (n = 7–8) and Cyp27b1 (n = 7) was also similar between Nuf/+ mice and WT mice. (h) Immunohistochemistry showed a marked increase in CLDN14 in both Nuf/+ and WT animals treated with DHT for 3 days. No apparent differences were observed between Nuf/+ and WT animals. Scale bar indicates 50 µm. *P < 0.05 for WT vs. Nuf/+ by Student's independent t-test (two-tailed) (b) or by repeated measurements two-way ANOVA with Bonferroni's post hoc test; interaction between time and genotype was significant (P = 0.0352) (c). Data are shown as mean ± SEM.

Experimental protocol 6

5 WT and 7 Nuf/+ female mice (age 8–11 weeks) were used in the experiment. To increase plasma Ca²⁺ towards normal levels in Nuf mice, only Nuf/+ mice were placed on 0.04% w/v DHT-containing diet for 7 days, whereas WT mice received standard diet. Mice were first maintained in regular cages, then placed in individual metabolic cages for acclimatization for 3 days, after which 24 h urine was collected on the 7th day. Mice were anesthetized and euthanized as above. Data are shown in Figure 4.

Figure 4.

The renal CaSR becomes activated after DHT treatment in Nuf/+ mice. (a) Plasma Ca²⁺ remained significantly lower in Nuf/+ mice treated with DHT for 7 days compared to untreated WT mice (n = 5–7). P = 0.0003. (b) Urinary Ca²⁺ excretion was significantly higher in Nuf/+ mice following DHT treatment compared to untreated WT mice (n = 5–7). P = 0.0311 (c) Fractional urinary excretion of Ca²⁺ (FE_Ca²⁺) was also significantly higher in DHT-treated Nuf mice than untreated WT mice (n = 5–7). P = 0.0133. (d) The expression of Cldn14 in the kidney was significantly higher in Nuf/+ mice compared to untreated WT animals (n = 5–7). P = 0.0139. (e) In addition, increased renal Cyp24a1 expression was present in Nuf/+ mice (n = 5–7). P < 0.0001. (f) In contrast, the expression of Cyp27b1 was significantly reduced in Nuf/+ mice (n = 5–7). P = 0.0004. (g) Immunohistochemistry showed a marked increase of CLDN14 in TAL of DHT-treated Nuf/+ mice, whereas this was absent in WT animals. Scale bar indicates 50 µm. Data are shown as mean ± SEM. *P < 0.05 by Student's independent t-test (two-tailed).

Experimental protocol 7

18 C57BL7/6J male mice (average age 10 weeks) were randomly divided into three groups with a similar average body weight. One group of animals served as controls, one group received infusion with 5 mg/ml etelcalcetide hydrochloride (Parsabiv; Amgen) via Alzet osmotic 2002 minipumps (Alzet, Cupertino, CA, USA) with a delivery rate of 0.5 µl/h (i.e. 2.5 µg/h), and one group received infusion with 20 mg/ml etelcalcetide hydrochloride via Alzet osmotic 2001 minipumps (Durect Corporation, Cupertino, CA, USA) with a delivery rate of 1.0 µl/h (e.g. 20 µg/h). Three days after insertion of minipumps, animals were housed in individual metabolic cages. After acclimatization for 3 days, baseline 24 h urine was collected on day 7. Mice were anesthetized and euthanized as above. Data are shown in Figure 5.

Figure 5.

Etelcalcetide hydrochloride leads to hypocalcemia but does not activate the renal CaSR. (a) WT mice infused with etelcalcetide hydrochloride showed a dose-dependent reduction in blood Ca²⁺ concentrations after 7 days of treatment (n = 6). P < 0.0001 for vehicle vs. 2.5 µg/h etelcalcetide hydrochloride and vehicle vs. 20 µg/h etelcalcetide hydrochloride; P = 0.0003 for 2.5 µg/h etelcalcetide hydrochloride vs. 20 µg/h etelcalcetide hydrochloride. (b) Plasma intact PTH was significantly lower in 2.5 µg/h etelcalcetide hydrochloride-treated mice than vehicle treated mice. P = 0.0314 (c) No statistically significant differences in urinary Ca²⁺ excretion were found between the two groups receiving etelcalcetide hydrochloride compared to vehicle-treated animals (n = 6). (d) No changes in renal Cldn14 were present between vehicle-treated animals and the two groups of animals that received a different dose of etelcalcetide hydrochloride (n = 4–6). (e) The renal expression of Cyp24a1 was significantly lower in animals having received 20 µg/h etelcalcetide hydrochloride compared to vehicle-treated animals (n = 4–6). P = 0.0064 (f) No significant differences were present for Cyp27b1 between any of the groups (n = 4–6). (e) CLDN14 was not observed by immunohistochemistry in mice treated with either vehicle, 2.5 µg/h etelcalcetide hydrochloride or 20 µg/h etelcalcetide hydrochloride. Scale bar indicates 50 µm. Data are shown as mean ± SEM (n = 4–6). *P < 0.05 by one-way ANOVA with Tukey's post hoc test.

Biochemical measurements

Blood ionized Ca²⁺ concentrations were measured by means of an ABL835 FLEX analyzer (Radiometer, Copenhagen, Denmark). Plasma was isolated from the remaining blood by centrifuging at 4500 g for 10 min at 4 °C and subsequently stored at −80 °C for further analyses. Urine Ca²⁺ concentrations were measured using a colorimetric assay.²⁷ Plasma and urine creatinine concentrations were measured using an ABX Pentra Creatinine 120 CP kit (HORIBA ABX SAS, Montpellier, France) on a Microlab 300 analyzer (ELITechGroup, Puteaux, France), according to the manufacturer's protocol. For experimental protocol 6 and 7, plasma and urine creatinine concentrations were measured by the Department of Clinical Biochemistry and Pharmacology of Odense University Hospital. Intact plasma parathyroid hormone (PTH) levels were determined with a mouse PTH 1–84 ELISA kit (Immutopics International, San Clemente, CA).⁸

Fractional excretion of Ca²⁺, the percentage of filtered Ca²⁺ that is excreted in the urine, was calculated as: $\frac{{\left[\text{Creatinine}\right]}_{\text{plasma}}\times {\left[\mathrm{C}{\mathrm{a}}^{2+}\right]}_{\text{urine}}}{{\left[\mathrm{C}{\mathrm{a}}^{2+}\right]}_{\text{ionized},\text{blood}}\times {\left[\text{Creatinine}\right]}_{\text{urine}}}\times 100\%$²⁸.

RNA extraction and semi-quantitative PCR

Kidneys were snap-frozen in liquid nitrogen after isolation and stored at −80 °C for future RNA extraction. RNA was extracted from the tissues using TRIzol (Invitrogen, Carlsbad, CA, USA) and processed as previously.^26,29 In brief, after DNAse treatment (Invitrogen, Carlsbad, CA, USA), cDNA synthesis was performed using the SensiFAST cDNA Synthesis Kit (Bioline, London, UK). Then, qPCR was performed in triplicate for each sample on the QuantStudio 6 Pro Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) using TaqMan Master Mix (Applied Biosystems, Waltham, MA, USA) and specific probes and primers (Table S1). Probes and primers were obtained from Applied Biosystems, Foster City, CA. For gene expression in the experiments employing the Nuf mice, analysis was performed using the 2^−ΔΔCt approach with 18S rRNA as a reference gene. For the experiments with etelcalcetide hydrochloride, Gapdh was used as a reference gene. This was because 18S showed significant differences between groups.

Immunohistochemistry

Immunohistochemistry was performed as described before.²⁸ Kidneys were immersion-fixed in 10% formalin overnight at 4 °C. They were then stored in PBS until paraffin embedding. Embedding was performed using a Tissue-Tec Vacuum Infiltration Processor 6 (Sakura Finetek, Torrance, CA, USA) and then kidneys were cut into 2 μm thick sections with a HM 355S Automatic Microtome (Thermo Scientific, Kalamazoo, MI, USA). For immunostaining, sections were processed with Tissue-Tek Tissue-clear (Sakura Finetek, Brøndby, Denmark) and a graded series of ethanol. Subsequently, they were boiled for 10 min in Tris-EGTA buffer (10 mM Tris, 0.5 mM EGTA, pH = 9.0). Endogenous peroxidase activity and free aldehyde groups were blocked by incubating in 0.6% H₂O₂ and 50 mM NH₄Cl in PBS. Then, sections were incubated overnight at 4 °C using a mouse monoclonal antibody against CLDN14 (Clone 5) produced in-house and extensively characterized⁹. After washing, sections were incubated with secondary HRP coupled antibodies followed by addition of DAB⁺ Substrate Chromogen System (Dako Omnis, Glostrup, Denmark). For each experiment, 6 mice were tested per genotype and representative images are shown. For the immunohistochemical analyses of the parathyroid and renal cortex, sections were processed as described above and stained using a rabbit polyclonal antibody against the CaSR (PB9924, Boster Biological Technology, Pleasanton, CA, USA).

Clinical data from ADH1 patient

Blood ionized Ca²⁺ and creatinine, urine Ca²⁺ and creatinine, and blood PTH were determined in the University of Alberta Hospital central laboratory by validated clinical biochemical protocols. Ionized blood Ca²⁺ concentrations between 1.09 and 1.25 mM are considered to be normal, while hypercalciuria is considered to be a urinary Ca²⁺/Cr > 0.6.³⁰ Normal plasma PTH is 1.4–6.8 pmol/L and normal serum creatinine is 50–120 μmol/L. We included data that had either paired blood ionized Ca²⁺ and parathyroid horomone levels or blood ionized Ca²⁺ and urine Ca²⁺ and creatinine levels. Due to the many samples collected over the life of this child, we averaged ionized Ca²⁺ and FECa or urine Ca²⁺/Cr over a range of blood ionized Ca²⁺ and plotted them in Figure 6.

Figure 6.

Ca²⁺ homeostasis in a patient with ADH1 (a) Treatment of a patient with ADH1 in order to normalize blood Ca²⁺ levels was associated with the occurrence of hypercalciuria when the patient was normocalcemic (n = 6–19 individual measurements per cluster). The vertical dashed red line represents the lower limit of normal for Ca²⁺ (ionized blood Ca²⁺ between 1.09 and 1.25 mM) and the horizontal red line represents the upper limit of urinary Ca²⁺ excretion considered normal (normal Ca²⁺/Cr is > 0.6). (b) In addition, fractional urinary excretion of Ca²⁺ (FE Ca²⁺) was increased after treatment leading to normocalcemia (n = 6–19 individual measurements per cluster). The horizontal dashed red line represents the lower limit of normal plasma Ca²⁺. (c) Plasma intact PTH levels (iPTH) were suppressed during all collection periods (n = 6–19 individual measurements per cluster). The horizontal dashed red line represents the lower limit of normal PTH for the clinical laboratory (Normal range is 1.4–6.8 pmol/L). d) Medullary calcinosis, a complication of treatment-induced hypercalciuria in ADH1, was observed by ultrasound.

Cldn14 reporter assay in HEK293 cells

Cloning of the Cldn14 promoter, which contains a CaSR-responsive element has been described previously.31, 32, 33 HEK293 cells were purchased from the ATCC (Rockville, MD, USA; RRID:CVCL_0045) and cultured in DMEM with 10% FBS and 5% penicillin, streptomycin and glutamine as previously.³⁴ Cells were transfected via the calcium-phosphate method with the reporter construct, the pRL-TK renilla luciferase vector as an internal control and varying concentration of a vector encoding myc-tagged CaSR (Origene Inc.). After 24 h incubation with standard medium containing 1.5 mM calcium, the cells were assayed for Cldn14 reporter expression using a luciferase assay system from Promega Corporation, according to the manufacturer's directions. Luciferase activity was assayed with 10 mL of extract on a DLR Ready, TD-20/20 Luminometer (Turner Design).

Western blotting

Samples were processed as described in detail previously.^9,35 Briefly, homogenized tissue was pelleted by centrifugation at 2000 g to remove nuclei and larger organelles and then at 17,000 g. Protein concentration was measured on the 17,000 g pellet using the DC protein assay (Bio-Rad, Copenhagen, Denmark) and samples loaded on polyacrylamide minigels (Bio-Rad). Samples were then electroeluted (25 mM Tris base, 192 mM glycine, 20% methanol, or 10% ethanol) to PVDF membranes (Merck, Darmstadt, Germany). Membranes were blocked in 5% skimmed milk in PBS with Tween (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, 0.05% Tween, pH 7.4) and incubated overnight at 4 °C with rabbit polyclonal antibodies against the CaSR (PB9924, Boster Biological Technology), mouse monoclonal anti-CLDN14 antibody⁹ or rabbit anti-β-actin antibody (Abcam, Cambridge, UK; #Ab8227; RRID:AB_306371). Following washing, membranes were incubated with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG1 gamma 1 heavy chain antibodies (Rockland Immunohistochemicals, Limerick, PA, USA; #610-103-040; RRID:AB 2614833) for CLDN14 or secondary anti-rabbit HRP conjugated antibodies for CaSR and β-actin (HRP, DakoCytomation) and visualized using an enhanced chemiluminescence system (ECL) on a Chemidoc XRS+ system (Bio-Rad). For each sample, the intensity ratio of protein of interest/β-actin was determined and normalized to the average of the control animals.

Lysate from the luciferase assay was subjected to SDS-PAGE and immunoblotting after mixing 3:1 with RIPA buffer as previously⁸. A mouse primary anti-CaSR monoclonal antibody (1:2000; GTX19347, GeneTex, Zeeland, MI, USA; RRID:AB_423536) or a mouse primary anti-Myc (9B11) monoclonal antibody (1:1000; #2276, Cell Signaling Technology, Danvers, MA, USA) was applied followed by a secondary horseradish peroxidase-coupled secondary antibody (1:5000, Santa Cruz Biotechnology, Santa Cruz, CA). This enabled detection of the CaSR at the appropriate molecular weight after incubation with Western Lightning Plus ECL reagents (PerkinElmer, Boston, MA) and visualized using a Kodak Image Station 440CF (Kodak, Rochester, NY).

Ethics

All animal experiments were conducted in accordance with Danish Law under animal experimental permits #2014-15-0201-00043 and #2019-15-0201-01629. With respect to the patient data, in consultation with the Research Ethics Board at the University of Alberta, formal ethical approval was not considered necessary, however, assent from the patient was provided as was formal written consent to include the clinical details of our subject by his guardian.

Statistical analyses

Sample size consisted of at least 5–6 animals per group and was based on our previous experience, as this was sufficient to previously find differences in renal Cldn14 expression;^8,9 each experimental unit refers to a single animal. No inclusion or exclusion criteria were applied; all allocated animals were used and no data was excluded from analysis. Randomization was applied where relevant (i.e. in the allocation of WT mice to treatment or control groups in experimental protocols 1 and 7) by giving the individual animals different identification numbers and manually assigning these over treatment groups. No blinding was performed. For the metabolic cage experiments, cages for the different groups were placed alternatively on the shelf. All data are available upon request. For all graphs, circles indicate female animals and squares male animals. Given the small sample size, we assumed that all data were normally distributed.

Statistical analyses were performed using GraphPad Prism 8.3.0 (GraphPad software, San Diego, CA, USA). Comparisons between more than two groups were performed using one-way ANOVA with Tukey's post hoc test or repeated measurements two-way ANOVA with Bonferroni's post hoc test, as indicated in the figure legends. Analyses between two groups were performed using Student's independent t-test (two-tailed). For all analyses, a P value < 0.05 was considered statistically significant. Data are presented as mean ± SEM

Role of the funding sources

The funding sources were not involved in the study design, data collection, data analyses, data interpretation, in the writing of the manuscript or in the decision to submit the manuscript for publication.

Results

Renal Cldn14 expression is dose-dependently regulated by blood Ca²⁺ levels via the CaSR

Previous studies found that CLDN14 is regulated by hypercalcemia and calcimimetics in vivo^8,9,11 and that Casr knockdown attenuates Ca²⁺-dependent increases in CLDN14 expression in vitro.¹¹ To assess whether Cldn14 expression increases in a dose dependent manner as plasma Ca²⁺ increases and thus can serve as a marker of kidney CaSR activity, we administered the synthetic vitamin D analog DHT at different dosages and durations to wildtype mice, as active vitamin D increases plasma Ca²⁺ by stimulating intestinal absorption, bone resorption and renal tubular Ca²⁺ reclamation.³⁶ Ca²⁺ levels were increased after a 3-day treatment with 0.4% w/v, 0.08% w/v and 0.04% w/v DHT compared to untreated animals, with the greatest increase in animals fed the highest dose of DHT (Figure 1a). Administration of 0.4% w/v DHT for 2 days, but not 1 day, also increased blood Ca²⁺ levels (Figure 1a). These changes were paralleled by proportionate increases in renal Cldn14 gene (Figure 1b) and CLDN14 protein expression (Figures 1c and S1a), with the highest dosage (0.4% w/v DHT) of DHT resulting in the greatest increase compared to control mice. Similarly, immunohistochemical analysis of kidney sections showed a strong CLDN14 signal in animals treated for 3 days with 0.4% w/v DHT, while CLDN14 expression was also clearly increased in animals treated with 0.08% w/v DHT for 3 days or 0.4% w/v DHT for 2 days. CLDN14 expression was largely absent in control mice (Figure 1d). Curve fitting using a power trend found correlation (R² = 0.73) between blood Ca²⁺ levels and Cldn14 gene expression, with the lowest ∆Ct values (and hence the highest expression) at the highest blood Ca²⁺ concentrations (Figure 1e).

To determine whether hypercalcemia drives Cldn14 expression via the CaSR, 0.4% w/v DHT was administered for 3 days to WT and kidney-specific Casr^−/− mice (KS-Casr^−/−). After DHT treatment, renal Cldn14 expression was significantly lower in kidney-specific Casr^−/- (KS-Casr^−/−) mice than in WT mice (Figure 1f). Similarly, both CLDN14 and CaSR were barely detectable by western blotting on kidney lysates from KS-Casr^−/− mice compared to WT mice after DHT treatment (Figures 1g and S1b,c). Marked CLDN14 expression was observed using immunohistochemistry in hypercalcemic WT animals, but not KS-Casr^−/− mice following DHT-treatment (Figure 1h).

Nuf mice are hypocalcemic with suppressed PTH but without altered renal Cldn14 expression

Nuf/+ and Nuf/Nuf mice, which have overactivating mutations in the Casr, were used to study the role of the renal CaSR in ADH1. These mice are hypocalcemic as reported before,²³ with Nuf/Nuf mice exhibiting more severe hypocalcemia than Nuf/+ mice (Figure 2a). Plasma PTH concentrations were reduced in Nuf/+ and Nuf/Nuf mice in comparison to WT mice (Figure 2b). Nuf/+ and Nuf/Nuf mice both had reduced urinary Ca²⁺/Cr relative to wild-type mice (Figure 2c). Despite the hypocalciuria, fractional urinary excretion of Ca²⁺ was unaltered (Figure 2d) as previously.³⁷

To investigate if the renal CaSR is activated in the Nuf mice, we measured expression of Cldn14. However, Cldn14 expression was unaltered in Nuf/+ and Nuf/Nuf mice compared to WT mice (Figure 2e). In addition, the renal expression of Cyp24a1 (Figure 2f) and Cyp27b1 (Figure 2g) was unchanged in Nuf mice. Immunohistochemistry showed low or absent expression of CLDN14 in all genotypes (Figure 2h). Given the similar phenotype of heterozygous and homozygous Nuf mice and that patients with ADH1 are heterozygous for activating mutations, we only used heterozygous Nuf mice for the remainder of experiments.

Activation of renal CaSRs is only achieved when blood Ca²⁺ levels are increased in Nuf mice

We next attempted to raise blood Ca²⁺ levels in Nuf/+ mice by administering a high (2% w/w) Ca²⁺ diet to WT and Nuf/+ mice for 6 days. However, as with ADH1 patients, this approach was unsuccessful at raising blood Ca²⁺ concentrations (Figure S2a). Urinary Ca²⁺/Cr remained lower in Nuf/+ mice than in WT animals (Figure S2b) with similar fractional urinary excretion (Figure S2c), although fractional excretion was markedly higher for both genotypes compared to a normal diet (Figure 1d). Renal Cldn14 expression was similar between the genotypes (Figure S2d) and so was Cyp24a1 (Figure S2e) and Cyp27b1 (Figure S2f). Immunohistochemistry revealed no CLDN14 expression in either Nuf/+ mice or WT mice (Figure 2e).

We next aimed to investigate the effects of hypercalcemia on Nuf mice by adding 0.4% w/v DHT to the food of WT and Nuf/+ mice for 3 days to increase blood Ca²⁺ levels.²⁵ The DHT-induced hypercalcemia was similar in both genotypes (Figure 3a). Plasma PTH concentrations remained slightly lower in Nuf/+ mice (Figure 3b). DHT increased urinary Ca²⁺ excretion in both genotypes (Figure 3C). Remarkably, the increase in urinary Ca²⁺ excretion following DHT supplementation was significantly delayed in Nuf/+ mice compared to WT animals (Figure 3c). On day 3, the fractional urinary excretion of Ca²⁺ was not significantly different between genotypes, indicating that the percentage of filtered Ca²⁺ that is excreted is unchanged in Nuf/+ mice (Figure 3d). Consistent with similar plasma Ca²⁺, the expression of Cldn14, Cyp24a1 and Cyp27b1 was similar between Nuf/+ mice and WT mice (Figure 3e–g). Dietary DHT supplementation increased CLDN14 expression in both Nuf/+ and WT mice, with no obvious differences between the genotypes (Figure 3h).

Differential activation of the renal CaSR after normalization of plasma Ca²⁺ levels in Nuf/+ mice

Since DHT treatment induced hypercalcemia in both genotypes (Figure 3a), we next attempted to normalize the blood Ca²⁺ of Nuf/+ mice to WT levels. Therefore, we administered 0.04% w/v DHT only to Nuf/+ mice. Blood Ca²⁺ concentrations remained significantly lower in DHT-treated Nuf/+ mice compared to untreated WT mice, although the difference between Nuf/+ mice and wild-type animals was smaller compared to previous experiments (Figure 4a; see also 2a and S2a). However, despite lower plasma Ca²⁺, urinary Ca²⁺ normalized to creatinine (Figure 4c) and fractional excretion of Ca²⁺ (Figure 4c) were significantly increased in Nuf/+ mice. Renal Cldn14 expression was significantly higher in Nuf/+ mice compared to WT mice (Figure 4d). Furthermore, Cyp24a1 expression was significantly higher in Nuf/+ mice (Figure 4e), while Cyp27b1 expression was significantly decreased in Nuf/+ mice (Figure 4f). CLDN14 immunolocalization was detectable in the TAL of Nuf/+ mice, but not in the kidneys of untreated WT mice (Figure 4g).

Infusion of etelcalcetide hydrochloride leads to PTH suppression without activating the renal CaSR

The calcimimetic etelcalcetide hydrochloride is a novel therapeutic for secondary hyperparathyroidism, suitable for infusion. Using this calcimimetic, we sought to observe differing thresholds of activation of the CaSR in the parathyroid versus the kidney.^21,22 Infusion of 2.5 µg/h and 20 µg/h drug resulted in a dose-dependent decrease in blood Ca²⁺ concentrations in WT mice (Figure 5a). Plasma PTH concentrations were significantly lower in mice treated with 2.5 µg/h etelcalcetide hydrochloride compared to vehicle-treated animals (Figure 5b). No differences were present in urinary Ca²⁺ excretion (Figure 5c), while Cldn14 expression was also unaltered by etelcalcetide hydrochloride (Figure 5d). The renal expression of Cyp24a1 was significantly decreased in the group that received 20 µg/h etelcalcetide hydrochloride compared to the vehicle-treated group (Figure 5e). No significant differences were present for Cyp27b1 (Figure 5f). CLDN14 expression was not detectable in any of the treatment groups (Figure 5g). This data is consistent with pharmacological activation of the CaSR in the parathyroid and not the kidney at this dosage, using etelcalcetide hydrochloride.

Relationship between urinary Ca²⁺ excretion and blood Ca²⁺ concentrations in a patient with ADH1

To translate our findings to patients with ADH1, urinary Ca²⁺ excretion over time was plotted relative to blood Ca²⁺ concentrations in a patient with ADH1. This now 18-year-old boy presented at 2 years of age with muscle spasms secondary to hypocalcemia and was then found to have low plasma PTH levels, without hypercalciuria. Sequencing of the CASR revealed a novel mutation (c.C2351T, p.A784V). Given ongoing clinical symptoms including muscle spasms, a failure to gain appropriate weight and height and peripheral paresthesia, he was treated initially with calcitriol and Ca²⁺ supplementation. Ongoing symptoms despite this therapy, as well as persistent hypocalcemia, led to the decision to start subcutaneous synthetic PTH 1–34 (3.5 units t.i.d., teriparatide; Forteo) treatment at the age of 10. He remains on this currently at a dose of 20 μg, t.i.d. Strikingly, hypercalciuria (Ca²⁺/Cr > 0.6)³⁰ was invariably present when he was normocalcemic (ionized blood Ca²⁺ between 1.09 and 1.25 mM) (Figure 6a). Moreover, when blood ionized Ca²⁺ concentrations were below the normal range he was much less likely to have hypercalciuria (Figure 6b). Fractional urinary excretion of Ca²⁺ was increased when blood Ca²⁺ concentrations were within the normal range, compared to most of the time when hypocalcemia was present (Figure 6b). Blood PTH concentrations were strongly suppressed regardless of blood Ca²⁺ levels (normal range 1.4–6.8 pmol/L) (Figure 6c). Unfortunately, he developed nephrocalcinosis at 9 years of age (Figure 6d). His renal function remains preserved with a plasma creatinine of 69 μM at last follow-up (normal range 50–120 μmol/L).

In vitro activation of the CaSR

Data from the Human Protein Atlas suggests that CaSR expression is higher in the parathyroid than in peripheral tissues, which could explain the different CaSR activation thresholds between the parathyroid and kidney.³⁸ To investigate this, we performed paired stainings of the CaSR in mouse kidney and parathyroid gland, which were in line with increased abundance of the CaSR in the parathyroid (Figure 7a) compared to renal cortical TAL (Figure 7b). To test the effect of increasing CaSR expression in the presence of a consistent extracellular Ca²⁺ concentration sufficient to activate the CaSR (1.5 mM Ca²⁺), a CaSR-responsive Cldn14 promoter reporter construct was transfected with increasing amounts of a myc-tagged CaSR into HEK293 cells.³³ Luminescence (and thus CaSR activity) significantly increased proportionately with increasing amounts of the CaSR construct transfected (Figure 7c). Western blotting confirmed increased CaSR expression after transfection with increased amount of construct (Figures 7d and S3).

Figure 7.

Relationship between CaSR abundance and CaSR activity. (a) Immunohistochemistry revealed a marked immunoreactivity of CaSR in the parathyroid of mice. (b) CaSR immunoreactivity was also detected in the thick ascending limb of Henle's loop in mouse kidney, albeit at a lower intensity. Scale bar indicates 50 µm. (c) After transfection of HEK293 cells with different amounts of a construct encoding the human CaSR (hCaSR) and a fixed amount of a CLDN14-promoter reporter construct, higher reporter signal was observed after transfection of cells with an increased amount of construct (n = 4 independent experiments). Data are shown as mean ± SEM. P < 0.0001 for 0.005 µg vs. 0.14 µg, 0.05 µg vs. 0.254 µg, 0.01 µg vs. 0.14 µg, 0.01 µg vs. 0.254 µg, 0.02 µg vs. 0.254 µg; P = 0.002 for 0.02 µg vs. 0.14 µg. *P < 0.05 by one-way ANOVA with Tukey's post hoc test. (d) Increased hCaSR expression was confirmed by western blotting with an anti-CaSR antibody. Representative immunohistochemical images and western blots are shown; arrows indicate dimeric CaSR signal. (e) Pooling blood Ca²⁺ concentrations and Cldn14 gene expression (∆Ct; Cldn14 Ct – 18S Ct) data from all experiments revealed a non-linear relationship, which could be described with a power trend line that fitted the data points equally well for both genotypes (R² = 0.44 for WT and 0.46 for Nuf/+ mice). A leftward shift was observed for Nuf/+ mice, indicating lower ∆Ct values for Cldn14 (and thereby higher expression) at the same blood Ca²⁺ levels as WT mice. (f) A similar non-linear relationship was present between plasma PTH concentrations and blood Ca²⁺ concentrations (R² = 0.82 for WT and 0.78 for Nuf/+), yet PTH was already markedly suppressed in the Nuf/+ mice.

Discussion

The CaSR plays important roles in maintaining systemic Ca²⁺ balance, exemplified by the marked perturbations observed in patients with ADH1 mutations in the CaSR. However, it remains incompletely understood why hypercalciuria is not always observed in patients with activating ADH1 mutations, which otherwise present with hypocalcemia due to CaSR-dependent suppression of PTH release. Using Cldn14 as a gauge of chronic CaSR activation, we examined renal CaSR activation in a mouse model of ADH1 and following pharmacological CaSR activation. Our data suggest that differential activation thresholds for parathyroid and renal CaSRs exist, where activation of the CaSR in the parathyroid reduces circulating Ca²⁺ concentrations, which are insufficient to activate renal CaSRs, even in the presence of some activating ADH1 mutations. This is based on the following observations: (i) CaSR-dependent increases in the expression of Cldn14 are absent in Nuf mice or following calcimimetic administration. (ii) Cldn14 expression only increases following elevations in plasma Ca²⁺ concentrations above baseline levels in both WT and Nuf mice. (iii) A patient with ADH1 displays persistently suppressed blood intact PTH levels under hypocalcemic conditions, while hypercalciuria occurred when treatment succeeded in increasing blood ionized Ca²⁺ levels.

The underlying mechanism mediating a differential response of the parathyroid CaSR versus epithelial CaSRs at a given plasma Ca²⁺ concentration remains to be elucidated in detail. Immunohistochemical staining performed on paired samples from mouse kidney and parathyroid suggests the CaSR is more highly expressed in the parathyroid relative to other tissues. This is supported by data from the Human Protein Atlas.³⁸ Therefore, pharmacological activation or gain-of-function mutations in the CaSR may lead to greater total activation of the CaSR in the parathyroid. Consistent with this, in vitro transcription of Cldn14 increased proportionately to the amount of CaSR expressed, at the same extracellular concentration of Ca²⁺. Furthermore, tissue-specific signaling pathways may exist, resulting in activation of the CaSR in the parathyroid and the kidney by different extracellular Ca²⁺ concentrations. In fact, the CaSR-dependent signaling in renal TAL segments may differ from that reported in the parathyroid.³⁹ Furthermore, heterocomplex formation with other receptors, variance in G protein abundance or identity and direct interactions with signaling proteins may differ between sites.⁴⁰

Importantly, when Cldn14 expression is plotted against blood Ca²⁺ concentrations for all Nuf/+ experiments, curve fitting using a power trend line resulted in a similar relationship for WT (R² = 0.44) and Nuf/+ mice (R² = 0.46) (Figure 7e). However, this relationship is shifted leftwards for Nuf/+ mice. This is consistent with the renal CaSR being inactive at baseline in Nuf/+ mice, then becoming activated at a lower blood Ca²⁺ concentration in Nuf/+ than WT mice. A similar relationship is present for both genotypes between plasma PTH concentrations and blood ionized Ca²⁺ concentrations (R² = 0.82 and 0.78 for WT and Nuf/+, respectively) (Figure 7f). However, plasma PTH was already decreased in Nuf/+ mice at baseline blood Ca²⁺ levels, indicating baseline parathyroid CaSR activation in Nuf/+ mice. Based on the seemingly differential activation of the parathyroid and kidney CaSR observed in this study, we propose that alterations in the activity of the CaSR in the parathyroid and subsequent modulation of PTH secretion is a key response to maintain normocalcemia. In contrast, activation of renal CaSRs might be limited to states of hypercalcemia where a reduction in PTH fails to reduce plasma Ca²⁺ levels. Accordingly, Pth^−/− mice, but not Casr^−/− Pth^−/− double knockout mice are still able to withstand hypercalcemic challenges.⁴¹

Nuf mice exhibit hypocalcemia and reduced plasma PTH concentrations similar to the ADH1 patients, reflecting activation of the parathyroid CaSR.¹³ Hypocalcemia is not compensated by increased tubular reclamation of Ca²⁺ as evidenced by unaltered FECa between Nuf and WT mice at baseline. Thus, Nuf mice are unable to compensate for hypocalcemia by increasing renal Ca²⁺ reabsorption. This may be explained by lower circulating levels of PTH in Nuf mice, which is a known potent stimulator of Ca²⁺ reabsorption in the TAL and DCT. Similar mechanisms are likely to explain the observations from mice treated with etelcalcetide hydrochloride. In these models, the filtered Ca²⁺ load is expected to be decreased due to lower blood Ca²⁺ levels. This likely contributes to the reduced urinary Ca²⁺ levels observed. Importantly, renal CaSRs are not activated in this chronic setting as indicated by unaltered Cldn14 expression and the absence of hypercalciuria as would be expected by the presence of a CaSR with an activating mutation in the kidney. A substantial fraction of ADH1 patients are not hypercalciuric, which could be explained by the absence of renal CaSR activation. Often such patients display hypercalciuria after treatment to correct symptomatic hypocalcemia.13, 14, 15, 16, 17, 18^,20 This increase in urinary Ca²⁺ excretion observed after treatment may occur following direct CaSR activation in the kidney resulting in tubular Ca²⁺ wasting, in combination with an increased filtered load of Ca²⁺ and lower PTH levels. More severe gain-of-function mutations in the CASR causes ADH1 with Bartter-like features.42, 43, 44 Accordingly, two mouse models with different gain-of-function mutations, based on human ADH1 mutations, displayed an increase in urinary Ca²⁺ excretion at baseline, despite having similar blood Ca²⁺ concentrations to the Nuf mice.44, 45, 46 These differences illustrate that the phenotype resulting from genetic overactivation of the CaSR can be more severe depending on the exact mutation, which would allow the renal CaSR to become activated at even lower circulating Ca²⁺ concentrations. Consistent with differential renal and parathyroid CaSR activation, etelcalcetide hydrochloride-treated mice are hypocalcemic without activation of the renal CaSRs, while another calcimimetic, cinacalcet, potently activates the renal CaSR.^8,9 Etelcalcetide binds to the extracellular domain of the CaSR and cinacalcet to the seven-transmembrane domain,⁴⁷ the latter allowing for more potent modulation of the CaSR. Furthermore, differences in potency and dosage may also account for these observed effects.

Caveats and limitations

This study is not without limitations. First, Cldn14 is used as an indirect marker of CaSR activity in kidney. While PTH has been considered a valid marker of parathyroid function and CaSR activity, the use of Cldn14 to gauge renal CaSR activity is novel. Although we cannot fully exclude that other factors may also affect Cldn14 expression, we clearly show that increases in blood Ca²⁺ drive expression of Cldn14 and that this occurs via a CaSR-dependent process. This is consistent with our previous data showing that chronic hypercalcemia or CaSR activation increases renal Cldn14 expression by some 10–170 times above baseline resulting in marked alterations in CLDN14 protein levels.^8,9 This marked increase in CLDN14 expression occurs irrespective of whether hypercalcemia is induced by active vitamin D derivatives or via PTH.^8,9 The only hypercalcemic model presented to date with no significant changes in CLDN14 expression in the chronic setting, is the transgenic model presented in this manuscript, where the CaSR is deleted specifically in the kidney. Furthermore, we have identified a Cldn14 promoter element that is directly regulated by the CaSR when the receptor is activated.31, 32, 33 Collectively, these findings are in line with Cldn14 being a reasonable indicator of CaSR activity in the kidney. One should also consider that CaSR-dependent signaling in the TAL may differ from that reported in the parathyroid³⁹ so other surrogate markers for CaSR activity such as intracellular Ca²⁺ release, would not necessarily be a better indicator. The absence of changes in Cldn14 expression in Nuf mice at baseline is consistent with the CaSR not being activated in the kidney in these animals at baseline. While we cannot fully exclude that a yet-to-be-identified putative negative regulator of Cldn14 exists, raising Ca²⁺ in the Nuf mice, even to levels lower than those observed in wildtype mice resulted in predictable increases in Cldn14 expression. This is in line with renal CaSR activation only occurring when plasma Ca²⁺ levels are sufficiently high. Potentially hypocalcemia itself could drive compensatory mechanisms to increase tubular Ca²⁺ reclamation. However, this has not been described, aside from the well-known effects of PTH pathway activation, which is markedly reduced in these animals.

Nuf mice have a missense mutation in the fourth transmembrane domain of the CaSR,²³ while transmembrane mutations in human ADH1 patients, including the one described here, are usually present in transmembrane domains 5 and 6.¹³ Nevertheless, the Nuf mice show classical ADH1 features with hypocalcemia and inappropriately low PTH levels suggesting that this is a reasonable mouse model to study ADH1. As outlined above, varying severity of the gain-of-function ADH1 mutations have been noted, with a few patients presenting with hypercalciuria even when hypocalcemic, which is different from the mouse model and patient investigated in our studies. However, other animal models with more severe gain-of-function ADH1 mutations have been generated to reflect this and highlights the fact that some mutations give rise to hypercalciuria at baseline, even though the animals show comparable blood Ca²⁺ concentrations to those found in the Nuf mice.44, 45, 46 A limitation of our study is thus the use of a single genetic model for the mouse experiments and a single patient as an example case.

In conclusion, we demonstrate that renal CaSRs have different activation thresholds from the parathyroid and document an intricate interplay between the parathyroid and renal CaSRs in maintaining Ca²⁺ balance. These findings could aid in explaining why hypercalciuria, a significant risk factor for nephrocalcinosis and chronic kidney disease, is not consistently observed in patients with ADH1 as this may depend on ambient levels of Ca²⁺ in the blood of the patient and the severity of the ADH1 mutation. Furthermore, these findings may aid in the clinical management of patients to avoid aberrant renal CaSR activation when possible.

Declaration of interests

RTA has received consulting fees and a grant from Ardylex Inc to determine the mechanism of action of their drug Tenapanor, which is not related to this work. RTA has received honoraria for Ultragenyx for a lecture on FGF23 lowering therapies. WvM obtained an Erasmus+ grant for a graduate research stay in the lab of H. Dimke. The authors have declared that no conflict of interest exists.