Enhanced passive Ca2+ reabsorption and reduced Mg2+ channel abundance explains thiazide-induced hypocalciuria and hypomagnesemia (original) (raw)

The present study demonstrates that enhanced passive Ca2+ transport in the proximal tubule due to ECV contraction explains the hypocalciuria that develops during chronic thiazide treatment. Of crucial importance is that micropuncture experiments in HCTZ-treated mice demonstrated increased reabsorption of Na+ and Ca2+ in the proximal tubule, whereas Ca2+ reabsorption in the distal convolution accessible to micropuncture, which includes the site of Trpv5-mediated active Ca2+ reabsorption, was unaffected. Furthermore, we showed that thiazide-induced hypocalciuria can be elicited in Trpv5–/– mice, in which active Ca2+ reabsorption is abolished. In line with these data, we showed that the hypocalciuric response to thiazides parallels a compensatory increase in renal Na+ reabsorption secondary to an initial natriuresis. Chronic thiazide administration enhanced Mg2+ excretion and specifically reduced renal expression levels of the epithelial Mg2+ channel Trpm6. In addition, Trpm6 expression was severely decreased in NCC–/– mice. Therefore, the pathogenesis of hypomagnesemia in chronic thiazide treatment as well as Gitelman syndrome appears to involve Trpm6 downregulation.

ECV contraction is a known stimulus for paracellular Ca2+ reabsorption in the proximal tubule (2, 14). Hypovolemia triggers a compensatory increase of proximal Na+ reabsorption, which in turn enhances the electrochemical gradient driving passive Ca2+ transport in proximal tubular segments (1, 2, 14). HCTZ-treated mice exhibited a significant increase in hematocrit and concomitant decrease in body weight compared with nontreated controls, which confirms that ECV contraction occurred. Importantly, micropuncture experiments showed that HCTZ increases Ca2+ reabsorption in the proximal tubule, which results in decreased Ca2+ delivery to the last surface loop of the proximal tubule and to the first surface loop of the distal convolution. This was associated with an increased fractional Na+ reabsorption rate in the proximal tubule, as further substantiated by the reduced CLi in HCTZ-treated mice. In line with these data, chronic HCTZ treatment upregulated expression of NHE3, which is responsible for the majority of Na+ reabsorption in proximal tubules and thus provides the main driving force for passive Ca2+ reabsorption. Alternatively, thiazide-induced hypocalciuria has been previously attributed to increased active Ca2+ reabsorption. However, the in vivo micropuncture experiments indicated similar or even reduced absolute Ca2+ reabsorption along the accessible distal convolution of HCTZ-treated versus control mice, whereas Ca2+ delivery to the early accessible distal convolution was decreased by HCTZ treatment. Previous micropuncture studies in distal convolutions of mice lacking Trpv5 revealed that Ca2+ reabsorption is normal up to the early accessible sites but becomes abolished along the downstream accessible segments of distal convolution (24). This indicated that no Trpv5-mediated Ca2+ reabsorption occurs upstream to the early sites accessible to micropuncture, which is consistent with the localization of Trpv5 in DCT2 and the CNT (17, 24). Thus, the reduced delivery of Ca2+ to the early accessible distal convolution in HCTZ-treated mice cannot be the result of Trpv5-mediated active Ca2+ reabsorption. In accordance with these findings, chronic HCTZ administration still decreased urinary Ca2+ excretion in mice lacking Trpv5, in which active Ca2+ transport in the distal convolution is effectively abolished (24). This is further exemplified by the severe downregulation of the remaining Ca2+ transport proteins in Trpv5–/– mice, which was unaffected by HCTZ treatment. Together with the micropuncture data, these results indicate that the Ca2+-sparing effect of thiazides can be explained neither by increased Ca2+ entry through Trpv5 nor by enhanced active Ca2+ transport in general. Additional experiments investigating the time dependency of these HCTZ effects demonstrated that Ca2+ excretion is unaltered during the natriuretic response following HCTZ administration, again indicating that direct stimulation of active Ca2+ reabsorption by inhibition of NCC does not occur. In contrast, a profound decrease in urinary Ca2+ excretion followed the initial natriuresis and, importantly, paralleled a reduced net Na+ excretion. In line with the present data, we previously demonstrated that volume contraction mimics thiazide-induced hypocalciuria and volume repletion completely reverses this hypocalciuria in rats (7). In conclusion, these data demonstrate that enhanced proximal tubular Na+ transport, as a consequence of ECV contraction, stimulates paracellular Ca2+ transport and constitutes the molecular mechanism underlying thiazide-induced hypocalciuria.

Gitelman syndrome is an autosomal recessive disorder caused by loss-of-function mutations in the gene encoding NCC, with a phenotype resembling chronic thiazide administration (4, 11). Loffing et al. recently demonstrated that renal Trpv5 and NCX1 expression are unaffected in NCC–/– mice, which display hypocalciuria (25, 27). Accordingly, micropuncture experiments in NCC–/– mice showed that active Ca2+ reabsorption is unaltered in distal convolution but that fractional absorption of both Na+ and Ca2+ in the proximal tubule is increased (27). The latter observation is in line with the enhancement of passive Ca2+ reabsorption through increased Na+ reabsorption in the proximal tubule, with results similar to the micropuncture data obtained during chronic HCTZ treatment in the present study. Taken together, our studies demonstrate that increased passive Ca2+ reabsorption in the proximal tubule explains the Ca2+-sparing effect of both NCC inhibition and inactivation.

Previously, we showed that thiazide-induced hypocalciuria occurs in spite of reduced renal expression of Ca2+ transport proteins in the rat (7). In these experiments a significantly higher dose of HCTZ was administered compared with that in the present study. Loffing et al. previously showed structural damage to and loss of DCT cells during thiazide administration with equivalently high doses, and, therefore, we hypothesized that apoptosis explains the reduced expression levels (7, 26). However, this was never observed in thiazide-treated mice. In the past, we have tried to reproduce the observed apoptotic changes in mice using several mouse strains, both sexes of mice, and different thiazide diuretics and application protocols but were never able to produce any DCT cell apoptosis in mice (J. Loffing, unpublished observations). Accordingly, deleterious effects were not detected in the present study, and the increased NCC expression in HCTZ-treated animals points to the viability of the DCT (28, 29). However, both studies consistently show that thiazide-induced hypocalciuria persists despite the absence or reduced abundance of proteins essentially responsible for active Ca2+ transport.

Interestingly, Trpv5–/– mice displayed significant polyuria and acidic urinary pH, which facilitate the excretion of large quantities of Ca2+ by reducing the risk of Ca2+ precipitations (24, 30–32). HCTZ administration increased diuresis in wild-type animals but did not alter urine volume in Trpv5–/– mice. Furthermore, HCTZ normalized the acidic urinary pH in Trpv5–/– mice. It has been demonstrated that a high luminal Ca2+ concentration, by activating the Ca2+-sensing receptor in the apical membrane of the collecting duct, blunts water permeability through aquaporin-2 (33, 34). Therefore, the Ca2+-sparing effect of HCTZ might counteract the hypercalciuria-induced polyuria and remove the necessity for urine acidification in Trpv5–/– mice, which would suggest that the distal Ca2+ load directly influences urine volume and acidification in an effort to prevent kidney stone formation.

Hypomagnesemia, as a side-effect of chronic thiazide administration and a defining feature of Gitelman syndrome, remains unexplained. Chronic HCTZ administration increased urinary Mg2+ excretion in the presence of reduced serum Mg2+ levels while GFR remained unaffected, which confirms that the hypomagnesemia is due to renal Mg2+ wasting. Importantly, renal Trpm6 expression was reduced in HCTZ-treated animals, while NCC expression was enhanced, which illustrates that Trpm6 downregulation is a specific nondeleterious effect. Trpm6 constitutes a Mg2+-permeable channel localized along the apical membrane of the DCT to which active Mg2+ reabsorption is restricted (21–23). Mutations in Trpm6 were shown to be associated with renal Mg2+ wasting (21, 22). Furthermore, we previously demonstrated a similarly reduced Trpm6 expression in tacrolimus-induced hypomagnesemia (35). Thus, the present data suggested that chronic thiazide treatment results in a similar defect in active Mg2+ reabsorption. As Trpm6 and NCC exactly colocalize in the DCT, a direct inhibitory effect of decreased NaCl influx on active Mg2+ transport could in principle be involved (23). However, whereas Ca2+ reabsorption was diminished upon a single dose of HCTZ, urinary Mg2+ excretion remained unaltered within 24 hours after HCTZ administration, which contradicts the hypothesis that Mg2+ reabsorption is directly inhibited. This dissociation of Ca2+ and Mg2+ excretion can be explained when the relative contribution of the proximal tubule and the TAL in the passive reabsorption of these divalents is taken into account. The majority of Mg2+ is reabsorbed in the TAL due to the driving force generated by NKCC2-mediated NaCl reabsorption, while micropuncture studies have shown that thiazides particularly enhance proximal tubule Na+ transport (14, 36). Accordingly, we showed that HCTZ increases proximal tubular NHE3 expression without affecting the NKCC2 expression. Hypothetically, an additional defect of Mg2+ reabsorption in TAL could explain the increased Mg2+ excretion in HCTZ treatment. However, inborn as well as acquired defects in TAL Mg2+ reabsorption are consistently accompanied by hypercalciuria, which renders this hypothesis unfeasible when the Ca2+-sparing effect of thiazides is taken into account (37–40). In line with the data obtained during HCTZ administration, NCC–/– mice were also shown to display hypomagnesemia in the absence of hypokalemia (4, 25). In the present study we report that, similar to what occurs during chronic thiazide administration, renal Trpm6 abundance in the DCT is significantly reduced in NCC–/– mice. Therefore, Trpm6 downregulation may represent a general mechanism involved in the pathogenesis of hypomagnesemia in both Gitelman syndrome and chronic thiazide administration.

_NCC–/–_mice were previously shown to display widespread atrophy of DCT cells (27). This marked reduction in DCT plasma membrane area and, thereby, apical expression of Mg2+ channels could explain the observed renal Mg2+ wasting. As discussed above, deleterious effects of chronic thiazide administration particularly on DCT1 were previously suggested but did not occur in the present study (7, 26). Thus, at present it is unknown which mechanism is responsible for the observed Trpm6 downregulation. In this respect, it is interesting to note that aldosterone excess has been shown to be associated with renal Mg2+ wasting, whereas hypermagnesemia may accompany aldosterone deficiency (4, 41). In addition, it has been shown that the mineralocorticoid receptor antagonist spironolactone reduces urinary Mg2+ excretion in patients with Gitelman syndrome (42, 43). During thiazide-induced ECV contraction as well as in NCC–/– mice, aldosterone levels are increased (25). Therefore, hyperaldosteronism might hypothetically downregulate Trpm6 expression and, thereby, result in renal Mg2+ wasting.

In conclusion, this study addressing the mechanism underlying thiazide-induced hypocalciuria answers a long-standing question in physiology and medicine. Furthermore, it offers new insights concerning the hypomagnesemia accompanying chronic thiazide treatment and Gitelman syndrome.