Calcium-sensing receptor and aquaporin 2 interplay in hypercalciuria-associated renal concentrating defect in humans. An in vivo and in vitro study

Abstract

One mechanism proposed for reducing the risk of calcium renal stones is activation of the calcium-sensing receptor (CaR) on the apical membranes of collecting duct principal cells by high luminal calcium. This would reduce the abundance of aquaporin-2 (AQP2) and in turn the rate of water reabsorption. While evidence in cells and in hypercalciuric animal models supports this hypothesis, the relevance of the interplay between the CaR and AQP2 in humans is not clear. This paper reports for the first time a detailed correlation between urinary AQP2 excretion under acute vasopressin action (DDAVP treatment) in hypercalciuric subjects and in parallel analyzes AQP2-CaR crosstalk in a mouse collecting duct cell line (MCD4) expressing endogenous and functional CaR. In normocalciurics, DDAVP administration resulted in a significant increase in AQP2 excretion paralleled by an increase in urinary osmolality indicating a physiological response to DDAVP. In contrast, in hypercalciurics, baseline AQP2 excretion was high and did not significantly increase after DDAVP. Moreover DDAVP treatment was accompanied by a less pronounced increase in urinary osmolality. These data indicate reduced urinary concentrating ability in response to vasopressin in hypercalciurics. Consistent with these results, biotinylation experiments in MCD4 cells revealed that membrane AQP2 expression in unstimulated cells exposed to CaR agonists was higher than in control cells and did not increase significantly in response to short term exposure to forskolin (FK). Interestingly, we found that CaR activation by specific agonists reduced the increase in cAMP and prevented any reduction in Rho activity in response to FK, two crucial pathways for AQP2 translocation. These data support the hypothesis that CaR-AQP2 interplay represents an internal renal defense to mitigate the effects of hypercalciuria on the risk of calcium precipitation during antidiuresis. This mechanism and possibly reduced medulla tonicity may explain the lower concentrating ability observed in hypercalciuric patients.

Figure 1. AQP2 excretion during urinary concentration test.

For the urinary concentration test, urinary AQP2 excretion and osmolality were measured on samples obtained hourly after 20 µg intranasal DDAVP administration. AQP2 urinary excretion. (A) In normocalciuric children, DDAVP administration resulted in a significant increase in urinary AQP2 excretion. Hypercalciuric children had a high basal AQP2 excretion and DDAVP administration did not result in a significant increase in urinary AQP2 excretion. Data were analyzed by Wilcoxon Signed Rank ANOVA test for paired non parametric data (*p = 0.0023). (B) Comparison of basal AQP2 excretion in hypercalciurics and normocalciurics. patients. In hypercalciurics a significant higher AQP2 excretion was observed (p = 0.036 ANOVA test).

Figure 2. Urinary osmolality during DDAVP test.

(A) Baseline and maximal urine osmolality after DDAVP test in normocalciuric and hypercalciuric patients. (B) Urinary osmolality in samples collected hourly after DDAVP administration. In hypercalciurics, DDAVP treatment was accompanied by a significantly lower increase in urinary osmolality, indicating a reduced urinary concentrating ability in response to vasopressin in hypercalciuric subjects. The values obtained were compared by one-way Anova and Tukey's multiple comparison test (*P<0.005, **P<0.001, ***P<0.0001.).

Figure 3. Effect of CaR signaling on AQP2 trafficking in MCD4 cells.

Apical surface biotinylation. (A) MCD4 cells were preincubated with 5 mM Ca²⁺, 300 µM Gd³⁺ or 10 µM NPS-R 568 then exposed to FK10⁻⁴ M or left under control conditions. Apical membrane-expressed AQP2 was quantitated by apical surface biotinylation. FK-induced AQP2 membrane accumulation was significantly reduced in the presence of CaR agonists. CaR agonists induced a mild increase in AQP2 membrane expression even in the absence of FK stimulation. The total amount of AQP2 in the starting preparation was comparable in each experimental condition (total AQP2). (B) Densitometric analysis of the 29 kDa biotinylated AQP2 band. Results are expressed as mean values ± SEM. The values obtained in five independent experiments are expressed as percentages of the basal condition. Data were compared by one-way Anova and Tukey's multiple comparison test (* P<0.05 relative to ctr, # P<0.05 relative to FK).

Figure 4. Effect of ATP stimulation on AQP2 trafficking in MCD4 cells.

(A) MCD4 cells were preincubated with 100 µM ATP or used under control conditions and then stimulated with FK 10⁻⁴ M. The amount of apical AQP2 was quantitated by apical surface biotinylation. ATP caused AQP2 membrane accumulation comparable to that found in FK treated cells. The total amount of AQP2 in the starting preparation was comparable in each experimental condition (total AQP2). (B) Densitometric analysis of the 29 kDa biotinylated AQP2 band. Results are expressed as mean values ± S.E.M. The values obtained in three independent experiments are expressed as percentages of the basal condition. Data were compared by one-way Anova and Tukey's multiple comparison test (* P<0.05 relative to ctr.) (C) Immunolocalization of AQP2 and analysis by confocal microscopy. MCD4 cells were grown on permeable support to full confluence fixed and immunostained with antibodies against AQP2. In resting cells, AQP2 was mainly detectable in sub-apical vesicles (ctr). After FK treatment, AQP2 localized to the apical membrane (FK). A similar apical localization was observed in cells exposed to 5 mM calcium (Ca²⁺) or to NPS-R 568 (NPS R) or to ATP (ATP) treatments. By contrast, no AQP2 redistribution was observed after exposure to the inactive NPS enantiomer NPS-S (NPS-S).

Figure 5. RhoA activity in cells exposed to CaR agonists.

A. Scheme of the Raichu-RBD probe mechanism. Raichu-RBD contains YFP and CFP separated by rhotekin-RBD (RBD). Active RhoGTP binds RBD, separating the donor (CFP) from the acceptor (YFP) thus reducing FRET. As internal control MCD4 cells were incubated for 3 hours with C3 toxin (1 µg/ml), which inactivates Rho proteins, and FRET signals was recorded. As shown in the figure, C3 toxin (1 µg/ml for 3 hours) (n = 124) leaded to a ‘closed’ conformation increasing FRET compared to control cells (n = 71). B. Rho activity during CaR activation. RhoA activity was evaluated in MCD4 cells exposed to CaR agonists. MCD4 cells were preincubated with 5 mM Ca²⁺, 300 µM Gd³⁺ or 10 µM NPS-R 568 for 30 min and then stimulated with 10⁻⁴ M FK or analyzed at rest. The amount of active RhoA was evaluated by FRET using a probe consisting of a Rho-binding domain (RBD) of Rhotekin sandwiched by YFP and CFP (see Methods). In this system, any increase in RhoA activity results in a decrease in FRET efficiency. In non-treated cells, FK stimulation (n = 59) caused a significant decrease in the amount of active RhoA compared to control conditions (n = 51). In cells pretreated with 300 µM Gd³⁺, RhoA activity was significantly increased (decreased FRET signal, n = 50) compared to control untreated cells. The decrease in RhoA activity in response to FK was prevented in cells preincubated either with Gd³⁺ (n = 66) or NPS-R 568 (n = 60). Values are expressed as mean ± SEM Data were compared by one-way Anova and Tukey's multiple comparison test (*P<0.05 relative to control).

Figure 6. Actin cytoskeleton in cells exposed to CaR agonists.

A. Visualization of actin cytoskeleton. MCD4 cells were preincubated with 5 mM Ca²⁺, 300 µM Gd³⁺ or 10 µM NPS-R 568 for 30 min and then stimulated with 10-4 M FK or analyzed at rest. Cells were fixed in PFA and stained with phalloidin Alexa Fluor-555 to visualize F-actin. Confocal pictures were taken for each experimental condition. Experiments were performed three times with similar results. B. F-actin quantization by actin polymerization assay. Confluent cells were either left untreated (CTR) or stimulated with forskolin (FK) or treated as described above with CaR agonists. After staining with TRITC-phalloidin, cells were extracted with cold methanol and the fluorescence absorbance of extracts was read (540/565 nm). The values obtained were compared by a one-way Anova and Newman-Keuls multiple comparison test (#P<0.05).