Renal Ca²⁺ wasting, hyperabsorption, and reduced bone thickness in mice lacking TRPV5

Abstract

Ca²⁺ ions play a fundamental role in many cellular processes, and the extracellular concentration of Ca²⁺ is kept under strict control to allow the proper physiological functions to take place. The kidney, small intestine, and bone determine the Ca²⁺ flux to the extracellular Ca²⁺ pool in a concerted fashion. Transient receptor potential (TRP) cation channel subfamily V, members 5 and 6 (TRPV5 and TRPV6) have recently been postulated to be the molecular gatekeepers facilitating Ca²⁺ influx in these tissues and are members of the TRP family, which mediates diverse biological effects ranging from pain perception to male aggression. Genetic ablation of TRPV5 in the mouse allowed us to investigate the function of this novel Ca²⁺ channel in maintaining the Ca²⁺ balance. Here, we demonstrate that mice lacking TRPV5 display diminished active Ca²⁺ reabsorption despite enhanced vitamin D levels, causing severe hypercalciuria. In vivo micropuncture experiments demonstrated that Ca²⁺ reabsorption was malfunctioning within the early part of the distal convolution, exactly where TRPV5 is localized. In addition, compensatory hyperabsorption of dietary Ca²⁺ was measured in TRPV5 knockout mice. Furthermore, the knockout mice exhibited significant disturbances in bone structure, including reduced trabecular and cortical bone thickness. These data demonstrate the key function of TRPV5 in active Ca²⁺ reabsorption and its essential role in the Ca²⁺ homeostasis.

Introduction

Ca²⁺ is the most abundant cation in the human body and serves a number of important physiological functions, including fertilization, synaptic transmission, muscle contraction, blood clotting, and bone mineralization. The extracellular Ca²⁺ concentration is controlled by the kidney, intestine, and bone through the action of the calciotropic hormones, including parathyroid hormone (PTH) and 1,25-dihydroxycholecalciferol (1,25-(OH)₂D₃). In humans, the daily dietary Ca²⁺ intake is less than 1,000 mg, of which only 30% is absorbed in the intestinal tract. This percentage is significantly enhanced during growth, pregnancy, and lactation by increased levels of circulating 1,25-(OH)₂D₃. Although there is continuous turnover of bone mass, there is no net gain or loss of Ca²⁺ from bone in a young and healthy individual. This indicates that healthy adults excrete a maximum of 300 mg Ca²⁺ in the urine to balance the intestinal Ca²⁺ uptake and that the remaining 98% of the Ca²⁺ filtered in the glomeruli is reabsorbed along the nephron. The molecular mechanism responsible for Ca²⁺ absorption in the small intestine and the kidney was elusive for a long time.

The cloning of transient receptor potential (TRP) cation channel subfamily V, member 5 (TRPV5; originally called ECaC1) from vitamin D–responsive rabbit renal epithelial cells (1) and TRP cation channel subfamily V, member 6 (TRPV6; originally called CaT1) from rat duodenum (2) has ignited research into transcellular Ca²⁺ (re)absorption at the molecular level (1). Mammals harbor at least 21 genes of the so-called TRP channels, whose functions remain mostly unknown (3). TRPV5 has been implicated as the Ca²⁺ influx channel in the process of vitamin D–responsive active Ca²⁺ reabsorption in the kidney (1, 4). In comparison, the TRPV5 homolog TRPV6, which displays an amino acid sequence identity of about 75% to TRPV5, has been postulated to be the Ca²⁺ influx channel facilitating Ca²⁺ absorption in enterocytes (2). TRPV6 is ubiquitously expressed and has been implicated as part of the capacitative Ca²⁺ entry mechanism and, therefore, intracellular Ca²⁺ signaling (5). TRPV5 and TRPV6 are Ca²⁺-selective channels that belong, together with the temperature-activated vanilloid receptors, to the TRPV subfamily (3, 4). Genomic cloning has demonstrated that TRPV5 and TRPV6 form a unique pair of novel Ca²⁺ channels that are transcribed from distinct genes juxtaposed on human chromosome 7q35 (4).

The present study investigated the physiological function of TRPV5 by the generation and characterization of TRPV5 knockout mice. Here, we demonstrate that the new TRP member TRPV5 is the gatekeeper in active Ca²⁺ reabsorption in the kidney. Ablation of the TRPV5 gene seriously disturbs renal Ca²⁺ handling, resulting in compensatory intestinal hyperabsorption and bone abnormalities.

Methods

Generation of TRPV5-null mice.

The TRPV5 gene was cloned from a 129/Sv mouse genomic library. For TRPV5 gene information, see GenBank accession number AF336378. The following fragments were then subcloned into a modified lox-targeting vector containing three loxP sites (6). First, a 1.2-kb 5′ homology fragment was inserted upstream of the 5′ loxP site in the pAT-FRT- K13 plasmid (kindly provided by A. Trumpp, University of California at San Francisco, San Francisco, California, USA) using long-range PCR with Takara polymerase (Takara Biomedical, Tokyo, Japan). For 3′ homology, a PCR-amplified 4-kb fragment was subcloned in the PmeI site of the pBS KS(+) targeting vector adjacent to the negative selection gene HSV-Tk (kindly provided by F. Radtke, University of Lausanne). Subsequently, for the functional channel region, a 600-bp PCR fragment containing exon 13 encoding the entire pore-forming region of the TRPV5 protein was cloned into the NotI and AscI sites by exchange of the neomycin-flanked loxP cassette. Finally, these two vectors were combined to construct the neomycin-loxP–targeting vector. To this end, the XhoI–NotI fragment from the pAT-FRT-K13 plasmid, containing the 5′ homology, the frt-flanked neomycin cassette, and the 5′ loxP site, was cloned into the SalI and NotI site of the pBS KS(+) vector upstream of the vital region. Electroporation and selection of targeted clones has been described (7). A recombinant ES cell was used to generate chimeras, allowing germline transmission of the mutant allele (TRPV5^loxneo). Heterozygous mice (TRPV5^+/–) harboring a null allele were obtained by breeding TRPV5^loxneo mice with a germline EIIa-cre-deletor strain (8). Offspring of heterozygous null mutants were subsequently intercrossed to obtain the TRPV5^+/+, TRPV5^+/–, and TRPV5^–/– mice used in this study. The frt-flanked neomycin cassette was not deleted in these mice. In all experiments, B6.129TRPV5^–/–,N₂ mice were compared with their TRPV5^+/+ and TRPV5^+/– littermates. Adult mice 8–9 weeks old, fed ad libitum, were placed in metabolic cages for 48 hours (Techniplast, Buguggiate, Italy). Food and fluid intake was monitored by weighing “sipper” bottles before and after the test period. Mice were allowed to adapt to these cages for 1 day, and then 24-hour urine samples were collected. The animal ethics board of the University Medical Center Nijmegen approved all experimental procedures.

Genotyping.

Genotyping was confirmed by Southern blot analysis of tail-derived DNA digested with BamHI and hybridized to a 5′ probe that demonstrated successful generation of the TRPV5^loxneo and TRPV5^null alleles (Figure 1, a and b). Heterozygous mice harbor a band of 5.8 kb plus a 4.6-kb endogenous wild-type band. Subsequently, genotypes were determined by PCR with primers A (5′-GCCAAGTTCTAATTCCATCAGAAG-3′) and B (5′-CTCCTATTCCCTGAAACCGCA-3′) to detect the null allele; and C (5′-TTCCCCAACTGTTCTTCCTGG-3′) and D (5′-AGATCCTCCAGTGTCTCATGCT-3′) to detect the wild-type allele (Figure 1a). The presence or absence of a 500-bp fragment (null allele) or 483-bp fragment (wild-type allele) identifies animals with –/–, +/+, and +/– genotypes (Figure 1c).

Figure 1.

Creation of null mutant mice for the TRPV5 gene locus. (a) Targeted inactivation of the TRPV5 gene. Top, exon 13, encoding the pore-forming unit of TRPV5, is flanked by loxP sites. Middle, recombined TRPV5^loxneo locus. Bottom, targeted allele in which exon 13 is deleted by Cre recombinase. Black boxes indicate exons; probe and genotype primers (A–D) are depicted. neo^R, neomycin resistance cassette. (b) Representative Southern blot analysis of BamHI-digested tail-derived DNA isolated from TRPV5^+/+, TRPV5^+/–, and TRPV5^–/– mice. (c) Identification of mouse genotype by PCR analysis of tail-derived DNA. The PCR product in the upper gel shows the presence of the wild-type allele (+/+) using primers C and D; the lower gel shows the knockout allele (–/–) using primers A and B. Both alleles are detected in heterozygous animals (+/–). (d) TRPV5 (top images) and kallikrein (lower images) immunohistochemical costaining of kidney cortex from TRPV5^+/+ and TRPV5^–/– mice. (e) Immunopositive TRPV6 staining of renal distal tubules from TRPV5^+/+ and TRPV5^–/– mice. A color version of this figure is available online as Supplementary Figure 1 (http://www.jci.org/cgi/content/full/112/12/1906/DC1).

Serum and urine analysis.

Serum 1,25-(OH)₂D₃, PTH, osteocalcin concentrations, and alkaline phosphatase were assayed according standard procedures. Mouse serum PTH levels were measured using a kit from Immunotopics Inc. (San Clemente, California, USA). Deoxypyrillinodine was analyzed in urine (Biometra, San Francisco, USA). Electrolytes in urine were measured using an Autoanalyzer (Hitachi, Laval, Quebec, Canada).

Micropuncture.

Mice anesthetized by inactin/ketamine were prepared for micropuncture as described (9). For assessment of GFR of both kidneys and of single nephrons, [³H]inulin was infused intravenously. Urine was collected using a bladder catheter. The left kidney was prepared for micropuncture. On the kidney surface, the last loop of proximal tubules (LPTs) or the distal convolutions were identified and punctured for quantitative collection of tubular fluid. Tubular fluid volume was determined from column length in a constant bore capillary. The tubular fluid concentration of K⁺ was determined by a micro-flame photometer (Department of Pharmacology, University of Tübingen) (9) and of Ca²⁺, by a flow-through microfluorometer (NanoFlo; WPI, Sarasota, Florida, USA) using Fluo-3 (MoBiTec, Göttingen, Germany) for detection (10, 11).

Immunohistochemistry.

Antibodies against TRPV5 and TRPV6 were obtained by immunization of guinea pigs and rabbits, respectively, with channel-specific peptides as described previously (12). Kidneys from TRPV5^+/+ and TRPV5^–/– mice were cut into pieces, placed in 1% (weight/volume) periodate-lysine-paraformaldehyde fixative for 2 hours, and incubated overnight in phosphate-buffered saline containing 15% (weight/volume) sucrose. Subsequently, samples were frozen in liquid nitrogen and sections 5–7 μm in thickness were used for different staining procedures as described previously (12). Rabbit anti-kallikrein was purchased from Calbiochem (San Diego, California, USA). All negative controls, including sections incubated with preimmune serum, subjected to preabsorption for 1 hour with 10 μg/ml TRPV5 or TRPV6 peptide, or incubated solely with conjugated secondary antibodies, were devoid of any staining. Photographs were taken with a Bio-Rad MRC 1000 confocal laser-scanning microscope.

Quantitative real-time PCR analysis.

Total RNA was extracted from kidney and duodenal tissues using the TriZol Total RNA Isolation Reagent (Gibco BRL, Breda, The Netherlands). The RNA was subjected to DNAse treatment to prevent genomic DNA contamination. Thereafter, 2 μg RNA was reverse-transcribed by Moloney murine leukemia virus reverse transcriptase (Gibco BRL) as described previously (13). The cDNA obtained was used to determine TRPV6, calbindin-D_28K, calbindin-D_9K, and NCX1 mRNA expression levels, as well as mRNA levels of the housekeeping gene hypoxanthine-guanine phosphoribosyl transferase (HPRT) as an endogenous control, as described previously (13). Expression levels were quantified by real-time quantitative PCR on an ABI Prism 7700 Sequence Detection System (PE Biosystems, Rotkreuz, Switzerland).

In vivo ⁴⁵Ca²⁺ absorption assay.

Ca²⁺ absorption was assessed by measuring serum ⁴⁵Ca²⁺ at early time points after oral gavage. Mice were fasted 12 hours before the test. Animals were hemodynamically stable under anesthesia (urethane, 1.4 mg/g body weight) during the experiment. The solution used to measure Ca²⁺ absorption contained 0.1 mM CaCl₂, 125 mM NaCl, 17 mM Tris, and 1.8 g/l fructose and was enriched with 20 μCi ⁴⁵CaCl₂/ml (18 Ci/g; New England Nuclear, Newton, Massachusetts, USA). For the oral tests, 15 μl/g body weight of this solution was administrated by gavage as described previously (14). Blood samples were obtained at indicated time intervals (Figure 3c). Serum (10 μl) was analyzed by liquid scintillation counting. The change in the plasma Ca²⁺ concentration was calculated from the ⁴⁵Ca²⁺ content of the plasma samples and the specific activity of the administrated Ca²⁺.

Figure 3.

Renal and duodenal Ca²⁺ transport assays. (a) Fractional Ca²⁺ delivery to the LPT, to the distal convolution (DC), and to urine (U), measured by micropuncture experiments in TRPV5^+/+ and TRPV5^–/– mice. (b) Relation between K⁺ concentration in tubular fluid of the distal convolution and fractional Ca²⁺ delivery to these sites. Low K⁺ concentrations indicate early puncture sites and high K⁺ concentrations late aspects of distal convolution (2). Values are means ± SEM of six mice per group for urine data and of 19–24 nephrons per group for LPT and distal convolution collections. *P < 0.05 versus TRPV5^+/+. (c) Changes in serum Ca²⁺ (ΔμM) within 10 minutes of administration of ⁴⁵Ca²⁺ by oral gavage in TRPV5^+/+ and TRPV5^–/– mice (n = 12). Data are averaged values ± SEM from mice 8 weeks old. *P < 0.05, significant difference from wild-type mice. (d) Expression of calbindin-D_28K and NCX1 mRNA in kidney cortexes of TRPV5^+/+ and TRPV5^–/– mice (n = 9), assessed by quantitative real-time PCR analysis and calculated as a ratio to HPRT RNA. (e) Expression of TRPV6 and calbindin-D_9K mRNA in duodenums of TRPV5^+/+ and TRPV5^–/– mice, assessed by quantitative real-time PCR analysis and calculated as a ratio to the HPRT RNA (n = 9).

Bone analysis.

Femurs from 8- to 9-week-old TRPV5^+/+ mice (females, n = 9; males, n = 5) and TRPV5^–/– mice (females, n = 8; males, n = 5) were scanned using the SkyScan 1072 microtomograph (SkyScan, Antwerp, Belgium). Scans were processed and a three-dimensional morphometric analysis of the bone was performed, using free software of the 3D-Calculator project (http://www.eur.nl/fgg/orthopaedics/Downloads.html). Measured parameters were expressed according to bone histomorphometry nomenclature (15). In the femoral head, bone volume (BV), total bone marrow volume including trabeculae (TV), trabecular bone volume fraction (BV/TV), trabecular thickness (Tb.Th) (16), trabecular number (Tb.N) (16), connectivity density (CD) (measure of the interconnectivity of the trabecular network; ref 17), and structure model index (SMI) (predominance of the shape of the trabeculae; 0 = plate-like; 3 = rod-like; ref. 18) were determined. In the diaphysis, calculations were performed with regard to cortical volume (Ct.V), endocortical volume (Ec.V), total diaphyseal volume (Dp.V) (sum of Ct.V and Ec.V), cortical thickness (Ct.Th) and cortical bone volume fraction (Ct.V/Dp.V) (16), and moment of inertia (MOI) (measure of the geometrical distribution indicative of the mechanical strength per transversal cross-section). Dissected femurs were routinely processed for plastic embedding. “Plastified” bone sections were deacrylated and rehydrated. Tartrate-resistant acid phosphatase (TRAP) staining was used to specifically stain osteoclasts red, as described (19). The number and surface area of osteoclasts per bone surface area (N.Oc/BS and Oc.S/BS, respectively) were determined in the femoral metaphysis using the Bioquant Novoprime digital-imaging software (Bioquant, Nashville, Tennessee, USA).

Statistics.

Data are expressed as means ± SEM. Statistical comparisons were tested by one-way analysis of variance (ANOVA). P values less than 0.05 were considered statistically significant. All analyses were performed using the StatView Statistical Package software (Power PC version 4.51; Berkeley, California, USA) on a Macintosh computer.

Results and Discussion

Genetic ablation of TRPV5 in the mouse allowed us to investigate the requirement of TRPV5 functioning in maintaining the Ca²⁺ balance. The conditional targeting strategy was aimed at deleting the functionally critical region of the protein that encodes exon 13, comprising the entire pore region, using Cre-loxP–mediated recombination (Figure 1a). This procedure will allow the engineering of tissue-specific TRPV5 knockout mice strains in the near future, but in the present study TRPV5 null allele mutants were generated to address the overall function of this channel. To this end, TRPV5^loxneo/+ and EIIa-cre transgenic mice were crossed (8). The murine TRPV5 gene is encoded by 15 exons and spans about 15 kb (GenBank accession number, AF336378). Southern blot together with PCR analysis demonstrated the deletion of pore-encoding exon 13, as shown in Figure 1, b and c. Immunohistochemical costaining of kidney cortex using antibodies against TRPV5 and kallikrein, a specific maker for distal convolution that comprises distal convoluted and connecting tubules (20), confirmed that the mutant TRPV5 allele was inactivated (Figure 1d). Importantly, ablation of the TRPV5 locus did not affect renal TRPV6 expression, confirming the gene-specific inactivation of TRPV5 (Figure 1e).

Gross observation revealed no obvious phenotype in the TRPV5-null mice. TRPV5^–/– and TRPV5^+/– mice were fertile and had similar average litter sizes (Figure 2a). TRPV5^–/– pups nursed without aid and thrived to adulthood at a rate similar to that of wild-type mice, as assessed by size and weight at sexual maturity. TRPV5^–/– females displayed normal maternal behaviors such as crouching and nursing, and transfer of a nursing female to a new cage initiated nest-building activity and efficient pup retrieval.

Figure 2.

Phenotypic characterization of TRPV5 knockout mice. (a) Body weights at an age of 8 weeks of TRPV5^+/+, TRPV5^+/–, and TRPV5^–/– mice (n = 29–43 mice). (b) The 24-hour urine volumes obtained from mice maintained in metabolic cages (n = 9 mice, three mice per cage). (c) Analysis of pH of urine obtained from TRPV5^+/+, TRPV5^+/–, and TRPV5^–/– mice (n = 9 mice, three mice per cage). (d) Total excretion of Ca²⁺, Na⁺, and K⁺ in TRPV5^+/+, TRPV5^+/–, and TRPV5^–/– mice (n = 9 mice, three mice per cage). Of note, TRPV5^+/– mice were similar to TRPV5^+/+ mice for the parameters measured. Data are averaged values ± SEM from mice at an age of 8 weeks. *P < 0.05, significant difference from TRPV5^+/+ mice.

TRPV5 ^+/+, TRPV5 ^+/–, and TRPV5 ^–/– mice were housed for 48 hours in metabolic cages (three mice per cage), and urine and serum samples were collected to investigate whether the electrolyte metabolism was disturbed. Polyuria was consistently observed in TRPV5 ^–/– mice compared with heterozygous mutants and wild-type littermates (Figure 2b). Importantly, TRPV5^–/– mice displayed a significant calciuresis, as they excreted about six times more Ca²⁺ in their urine than did their TRPV5^+/+ and TRPV5^+/– littermates (Figure 2d, left panel). The urinary Ca²⁺ concentration of the knockout mice reached values as high as 20 mM, compared with 6 mM for TRPV5^+/+ littermates. In addition, urinary phosphate excretion was significantly increased in TRPV5^–/– mice compared with that of TRPV5^+/+ mice (512 ± 30 μmol and 327 ± 25 μmol per 24 hours per three mice, respectively; n = 6, P < 0.05). Polyuria facilitates the excretion of large quantities of Ca²⁺ by reducing the potential risk of Ca²⁺ precipitation. It has been demonstrated that a high luminal Ca²⁺ concentration activates the calcium-sensing receptor in the apical membrane of the inner medullary collecting duct. As a consequence, the arginine vasopressin–elicited (AVP-elicited) water permeability is blunted to prevent a further rise in urinary Ca²⁺ concentration and possibly stone formation (21). Hypercalciuria-induced polyuria has been observed in humans and animal models (22–24). Notably, TRPV5^–/– mice produced urine that was significantly more acidic than that of their TRPV5^+/+ and TRPV5^+/– littermates (Figure 2c). Acidification of the urine is known to prevent renal stone formation in hypercalciuria, as Ca²⁺ oxalate stones cannot be formed at a pH of 5–6 (25). Neutralization of the acidic urine obtained from TRPV5^–/– mice resulted in the development of large crystals, confirming that acidification of the urine of the knockout mice is crucial for the prevention of kidney stone formation. A better understanding of the molecular mechanism by which hypercalciuria occurs will be important for the development of clinical tools that may lead to screening of families of “stone formers” before the onset of overt disease and to early treatment.

A primary renal origin of the observed calciuresis was supported by the fact that hypercalciuria persisted when TRPV5^–/– mice consumed a Ca²⁺-deficient diet for 1 week (data not shown). Furthermore, the GFR was not different between TRPV5^–/– and TRPV5^+/+ mice (0.31 ± 0.01 ml/min and 0.30 ± 0.07 ml/min, respectively; n = 6), indicating a primary defect in renal Ca²⁺ reabsorption as the cause of calciuresis in TRPV5^–/– mice. To localize the site of defective Ca²⁺ reabsorption along the nephron, we performed micropuncture studies (9–11) in anesthetized TRPV5^–/– and TRPV5^+/+ mice (Figure 3a). Confirming the results of the experiments in the metabolic cages, TRPV5^–/– mice showed calciuresis as a consequence of reduced renal Ca²⁺ reabsorption. Quantitative free-flow collection of tubular fluid revealed unaffected Ca²⁺ reabsorption in TRPV5^–/– mice up to the LPT. In contrast, mean Ca²⁺ delivery to puncturing sites within the distal convolution was significantly enhanced in TRPV5^–/– mice.

Because K⁺ secretion occurs along the distal nephron sites accessible to micropuncture (together with water reabsorption in the connecting tubule and cortical collecting duct), the distal luminal [K⁺] was used as an indicator of the distal collection site (Figure 3b). Based on the shape of the relationship between distal luminal [K⁺] and fractional Ca²⁺ delivery, it is evident that, in TRPV5^–/– mice, in contrast to TRPV5^+/+ mice, fractional Ca²⁺ delivery increases with higher [K⁺], indicating a defect in Ca²⁺ reabsorption along the distal convolution, which is consistent with the expression site of TRPV5. Given the potential internephron variations, the observation that some nephrons of TRPV5^–/– mice with low luminal [K⁺] exhibit low and some high fractional Ca²⁺ delivery is consistent with a rapid increase in luminal Ca²⁺ delivery along the early distal convolutions of mice lacking TRPV5. The increase in mean fractional Ca²⁺ delivery with increasing luminal [K⁺] in TRPV5^–/– mice cannot be explained by a simple lack of Ca²⁺ reabsorption, but would be consistent with Ca²⁺ back-leak into the lumen in the absence of transcellular reabsorption. In comparison, if the primary Ca²⁺ reabsorption defect were upstream of the distal convolution, for example, in the thick limb of Henle’s loop, then fractional Ca²⁺ delivery should be enhanced in TRPV5^–/– mice distal nephrons irrespective of the luminal [K⁺]. This contrasts with the observation made in the TRPV5^–/– mice, indicating that the transport defect is localized primarily in the distal convolution. The fractional delivery of Na⁺ to the distal convolution was not significantly different between TRPV5^–/– and TRPV5^+/+ mice (4 ± 1% vs. 5 ± 1%, respectively; n = 6, P > 0.2), which argues against the idea that a primary decrease in Na⁺ reabsorption upstream of the distal convolution indirectly influences Ca²⁺ reabsorption.

In the Ca²⁺-reabsorbing epithelial cells of the distal convolution, the influx of Ca²⁺ is rapidly buffered by cytosolic calcium-binding proteins, that is, calbindin-D_28K and calbindin-D_9K, and the bound Ca²⁺ diffuses to the basolateral membrane where a Na⁺-Ca²⁺ exchanger (NCX1) and, to a lesser extent, an ATP-dependent Ca²⁺ pump extrude Ca²⁺ to the blood compartment, completing the process of transcellular Ca²⁺ transport (4). The expression of the Ca²⁺ transport proteins in the kidney was investigated by quantitative real-time PCR as described previously (13). Inactivation of the TRPV5 gene was accompanied by a decrease in calbindin-D_28K and NCX1 mRNA levels (Figure 3d). As transcellular Ca²⁺ reabsorption in the distal convolution was abolished in TRPV5^–/– mice, the concomitant decrease in calbindin-D_28K and NCX1 mRNA levels in the presence of elevated 1,25-(OH)₂D₃ levels suggests a regulatory mechanism controlled primarily by TRPV5. Thus, TRPV5 or the Ca²⁺ influx through TRPV5 possibly controls the transcription of the other Ca²⁺ transport genes, including calbindin-D_28K and NCX1. These findings emphasize the gatekeeper function of TRPV5. Although downregulation of renal calbindin-D_28K is secondary to TRPV5 ablation in the TRPV5^–/– mice, the reduced calbindin-D_28K level may further augment the severity of the hypercalciuria. Interestingly, calbindin-D_28K knockout mice fed a Ca²⁺ diet (1% weight/volume) displayed an approximately twofold increase in urinary Ca²⁺ excretion compared with that of their wild-type littermates (26), whereas TRPV5^–/– mice fed the same Ca²⁺ diet excreted about six times more Ca²⁺ in their urine than did their littermates.

In contrast to the excessive renal Ca²⁺ wasting noted, natriuresis and kaliuresis were not observed in the knockout mice, demonstrating the specific effect of TRPV5 inactivation on Ca²⁺ reabsorption (Figure 2d). Importantly, despite renal Ca²⁺ wasting, the serum Ca²⁺ concentration remained normal, suggesting successful compensatory mechanisms (Figure 4d). We investigated whether the impairment of transcellular Ca²⁺ reabsorption indirectly affected intestinal Ca²⁺ absorption. Active Ca²⁺ uptake by the small intestine is favored as compensation for renal Ca²⁺ wasting. To provide functional evidence for intestinal hyperabsorption in TRPV5^–/– mice, we assessed Ca²⁺ absorption by measuring serum ⁴⁵Ca²⁺ at early time points after oral gavage, as described (14). A significant increase in the rate ⁴⁵Ca²⁺ absorption was observed in TRPV5^–/– mice compared with their wild-type littermates, indicating a compensatory role for the small intestine (Figure 3c). Of note, this may actually represent an underestimation due to the applied single-⁴⁵Ca²⁺-isotope technique used instead of a dual-isotope technique (27). Whereas TRPV5 is expressed predominantly in the kidney, TRPV6 is highly expressed in small intestine and has been implicated as the Ca²⁺ influx mechanism in enterocytes (1, 2). To test this hypothesis, duodenal RNA was extracted from knockout and wild-type littermates and analyzed for expression of TRPV6 and calbindin-D_9K by quantitative real-time PCR analysis. TRPV6 and calbindin-D_9K expression levels were significantly upregulated in TRPV5^–/– mice, consistent with an increased Ca²⁺ absorption (Figure 3, c and e).

Figure 4.

Bone phenotypes of female and male TRPV5^+/+ and TRPV5^–/– mice. (a) Representative cross-sectional x-ray images of the femoral head (I), the lesser trochanter (II), and the diaphysis (III) in TRPV5^+/+ and TRPV5^–/– mice. Note the decreased cortical bone width in the trochanter and diaphysis (arrows). (b) MOI as a parameter of bone quality. (c) Three-dimensional reconstruction of femurs from TRPV5^+/+ and TRPV5^–/– mice. Note the reduced cortical and Tb.Th in TRPV5^–/– mice. (d) Changes in serum levels of 1,25-(OH)₂D₃, PTH, and Ca²⁺ in TRPV5^+/+, TRPV5^+/–, and TRPV5^–/– mice. Values are expressed as percentage of TRPV5^+/+ values ± SEM. The absolute values for PTH are TRPV5^+/+, 55 ± 15 pg/ml; TRPV5^+/–, 59 ± 14 pg/ml; and TRPV5^–/–, 60 ± 16 pg/ml; for 1,25-(OH)₂D₃, TRPV5^+/+, 239 ± 38 pmol/l; TRPV5^+/–, 225 ± 32 pmol/l; TRPV5^–/–, 691 ± 88 pmol/l; and for free Ca²⁺ concentration, TRPV5^+/+, 1.20 ± 0.02 mM; TRPV5^+/–, 1.25 ± 0.03 mM; and TRPV5^–/–, 1.20 ± 0.03 mM. *P < 0.05, significant difference from wild-type mice.

Serum 1,25-(OH)₂D₃ levels were significantly elevated in TRPV5^–/– mice (Figure 4d), which explains the compensatory increased intestinal TRPV6 and calbindin-D_9K mRNA expression and Ca²⁺ absorption. The TRPV5^–/– mouse is the first model to our knowledge to have sustained highly elevated 1,25-(OH)₂D₃ levels while remaining normocalcemic (Figure 4d) and normophosphatemic, as serum phosphate was not different between the TRPV5^+/+ and TRPV5^–/– mice (2.68 ± 0.29 and 3.04 ± 0.18 mM, respectively, n = 6, P > 0.2). The elevated 1,25-(OH)₂D₃ levels observed in vitamin D receptor (VDR)-null mice (14) and in rickets type II are normally accompanied by hypocalcemia and increased serum PTH levels. In contrast, in TRPV5^–/– mice with normal functioning VDRs, PTH was not significantly changed (Figure 4d). The elevated 1,25-(OH)₂D₃ levels in TRPV5^–/– mice were not due to increased expression of renal 25-hydroxyvitamin D₃-1α-hydroxylase mRNA, since no differences were measured in the expression of this enzyme between TRPV5^+/+ and TRPV5^–/– mice (100 ± 13% and 139 ± 29%, respectively; n = 4, P > 0.2). Studies to clarify the role of 1,25-(OH)₂D₃ in bone have been hampered by the complexity of distinguishing between disturbances in Ca²⁺ homeostasis and changes in serum 1,25-(OH)₂D₃ levels. Normal PTH plus normocalcemia make the TRPV5^–/– mouse a unique model to study the effect of 1,25-(OH)₂D₃ on bone.

The homeostasis of bone tissue is maintained by the balanced processes of bone resorption and formation. The bone formation marker alkaline phosphatase was not different between the groups (Table 1), while osteocalcin was slightly but significantly higher in sera of TRPV5^–/– versus TRPV5^+/+ mice. Bone resorption was studied by TRAP staining of femoral bone sections, by measurement of osteoclast number and surface, and by measurement of urinary deoxypyridinoline. Both N.Oc/BS and Oc.S/BS were significantly increased in female TRPV5^–/– mice, with males showing a similar trend (Table 1). Urinary deoxypyridinoline showed large variations and was not significantly different between TRPV5^–/– and TRPV5^+/+ mice (66 ± 8 nmol/nmol creatinine and 92 ± 21 nmol/nmol creatinine, respectively; n = 5–9, P > 0.2).

Table 1.

Bone analysis of TRPV5^+/+ and TRPV5^–/– mice

Detailed microcomputed tomography analyses of the femur demonstrated that Tb.Th in the femoral head of TRPV5^–/– mice was significantly reduced compared with that of TRPV5^+/+ mice (Table 1 and Figure 4, a and c). BV/TV was not different between the genotypes. This could not be explained by a difference in Tb.N (Table 1). Alternatively, it is possible that the trabeculae are longer, that is, they protrude further into the bone marrow cavity, and thereby compensate for reduced Tb.Th. CD and SMI were not altered in the femoral head, irrespective of phenotype and sex. Analyses of the diaphysis showed that Ct.V, Ct.V/Dp.V, and Ct.Th were significantly reduced in TRPV5^–/– versus TRPV5^+/+ mice (Table 1 and Figure 4, a and c). Finally, the polar MOI was significantly reduced in TRPV5^–/– mice, indicative of an altered geometrical distribution and reduced mechanical properties of the diaphyseal bone (Table 1 and Figure 4b). The observed changes in bone structure cannot be attributed to differences in body weight (Figure 2a). Also, the bone lengths of TRPV5^–/– and TRPV5^+/+ mice were similar. In summary, the unchanged alkaline phosphatase together with the elevated N.Oc/BS and Oc.S/BS in the TRPV5^–/– mice may explain the observed reduced trabecular and cortical thickness.

It is unclear whether the reduced Tb.Th and Ct.Th found in the TRPV5^–/– mice is a direct effect of the genotype or results from prolonged elevated 1,25-(OH)₂D₃ levels. Interestingly, a recent study of long-term treatment of mice with 1,25-(OH)₂D₃ demonstrated a decrease in cortical thickness and cross-sectional area and a 50% reduction in stiffness (28), which is in line with the current observations. We hypothesize that the bone phenotype of the TRPV5^–/– mice, due to prolonged elevated 1,25-(OH)₂D₃ levels, develops and progresses throughout life. However, at present it is premature to draw definitive conclusions about bone metabolism, and future experiments will certainly address this point in detail.

The epithelial Ca²⁺ channels TRPV5 and TRPV6 have been studied extensively in epithelial tissues controlling the Ca²⁺ homeostasis and exhibit a range of distinctive properties that distinguish them from other TRP channels (3, 4). TRP channels display an extraordinary assortment of selectivities and activation mechanisms, some of which represent previously unrecognized modes for regulating ion channels (3). Moreover, there is only limited information about the biological functions of TRP channels, which appear to be equally diverse and range from pain perception to male aggression. The present study demonstrated that the new TRP member TRPV5 is the gatekeeper in active Ca²⁺ reabsorption in the kidney. Ablation of the TRPV5 gene seriously disturbs renal Ca²⁺ handling, resulting in compensatory intestinal hyperabsorption and bone abnormalities. These deficiencies in Ca²⁺ handling have been reported frequently in patients with idiopathic hypercalciuria, although the molecular basis for this disorder remains unknown. Dysfunction of TRPV5 may contribute to such disturbances of Ca²⁺ homeostasis, and the TRPV5 knockout mouse is, therefore, a powerful animal model for elucidating the molecular mechanism underlying idiopathic hypercalciuria. Finally, the increased 1,25-(OH)₂D₃ levels and effects on bone structure indicate that alterations in TRPV5 may have implications for age-related bone disorders, including osteoporosis.