sign in sign up
<<
Home , Aquaporin, Aquaporin 2, Calcium channel, TRPV5, TRPV6

ARTICLES J Am Soc Nephrol 14: 2731–2740, 2003

Localization and Regulation of the Epithelial Ca2ϩ Channel TRPV6 in the

TOM NIJENHUIS, JOOST G. J. HOENDEROP, ANNEMIETE W. C. M. VAN DER KEMP, and RENE´ J. M. BINDELS Department of Physiology, Nijmegen Center for Molecular Life Sciences, University Medical Center Nijmegen, the Netherlands.

Abstract. The family of epithelial Ca2ϩ channels consists of with the other known Ca2ϩ transport , including two highly homologues members, TRPV5 and TRPV6, which TRPV5 and -D28K. Together, these data suggest a role 2ϩ 2ϩ 2ϩ constitute the apical Ca entry mechanism in active Ca for TRPV6 in 1,25(OH)2D3-stimulated Ca reabsorption in (re)absorption in kidney and . In kidney, TRPV5 these segments. Interestingly, distribution of TRPV6 extended expression has been extensively studied, whereas TRPV6 lo- to the CD, where it localized to the apical domain of principal calization and regulation has been largely confined to the small and intercalated cells, which are not generally implicated in intestine. The present study investigated the renal distribution active Ca2ϩ reabsorption. In addition, TRPV6 mRNA levels of TRPV6 and regulation by 1,25-dihydroxyvitamin D3 were quantified in a large set of tissues, and in the order of Ͼ (1,25(OH)2D3). In mouse kidney, TRPV6 was detected by decreasing expression level were detected: , at the apical domain of the distal con- Ͼ lung Ͼ , kidney, bone, cecum, heart Ͼ voluted tubules (DCT2), connecting tubules (CNT), and corti- colon Ͼ skeletal muscle Ͼ . Therefore, additional cal and medullary collecting ducts (CD). Furthermore, several physiologic functions for TRPV6 are feasible. In conclusion, putative –responsive elements were detected up- TRPV6 is expressed along the apical domain of DCT2, CNT, stream of the mouse TRPV6 start codon, and 1,25(OH)2D3 and CD, where TRPV6 expression is positively regulated by treatment significantly increased renal TRPV6 mRNA and 1,25(OH)2D3. expression. In DCT2 and CNT, TRPV6 co-localizes

Active Ca2ϩ absorption plays a key role in Ca2ϩ homeostasis Ca2ϩ-dependent inactivation (7,8). However, significant dif- and takes place in Ca2ϩ-transporting tissues, including kidney ferences exist in the N- and C-termini of TRPV5 and TRPV6, and intestine. In the mammalian genome, a distinct family of which may account for distinct functional and regulatory epithelial Ca2ϩ channels was recently identified, which pro- features. vides the molecular identity of the apical entry mechanism, In the kidney, active transcellular reabsorption of Ca2ϩ takes facilitating the active Ca2ϩ transport process (1). This family place in the (DCT) and the connecting of highly Ca2ϩ-selective channels is restricted to two members, tubule (CNT) and constitutes the fine-tuning mechanism de- TRPV5 (previously ECaC1), which was originally cloned from termining net urinary Ca2ϩ excretion (1,9,10). The renal ex- rabbit kidney, and TRPV6 (previously CaT1/ECaC2), which pression and regulation of TRPV5 has been extensively stud- was identified from rat duodenum (2,3). Genomic analysis ied. Briefly, TRPV5 is localized along the apical membrane of revealed that two , juxtaposed on human DCT2 and CNT and is regulated by vitamin D (1,25(OH)2D3), 7q35 and on mouse chromosome 6, encode these highly ho- dietary Ca2ϩ, and , substantiating the role of TRPV5 mologous but distinct epithelial Ca2ϩ channels (4–6). In ad- in renal active Ca2ϩ transport (11–13). Data on the localization dition to the conserved pore region, these channels share sev- and regulation of TRPV6 have been largely confined to the eral functional properties, including the permeation profile for small intestine, where the expression and regulation by 2ϩ monovalent and divalent cations, the high Ca selectivity, and 1,25(OH)2D3 strongly supports the involvement of this channel in the absorption of dietary Ca2ϩ (14,15). Therefore, TRPV5 is generally suggested to be the epithelial Ca2ϩ channel primarily Received May 9, 2003. Accepted August 5, 2003. responsible for renal transcellular Ca2ϩ transport, whereas Correspondence to Dr. René J.M. Bindels, 160 Physiology, University TRPV6 would serve as the apical Ca2ϩ entry mechanism in the Medical Center Nijmegen, P.O. Box 9101, NL-6500 HB Nijmegen, the Neth- erlands. Phone: 31-24-3614211; Fax: 31-24-3616413; E-mail: small intestine. [email protected] Interestingly, TRPV6 mRNA has been detected in the kid- 1046-6673/1411-2731 ney, but little is known about the localization and regulation of Journal of the American Society of Nephrology TRPV6 in this organ (16). The elucidation of the precise Copyright © 2003 by the American Society of Nephrology distribution of TRPV6 is crucial in understanding its DOI: 10.1097/01.ASN.0000094081.78893.E8 physiologic role in the kidney in general and particularly in 2732 Journal of the American Society of Nephrology J Am Soc Nephrol 14: 2731–2740, 2003 evaluating the potential contribution of TRPV6 to renal trans- structed. To this end, C57BL6 mice were sacrificed and kidney, bone, cellular Ca2ϩ reabsorption. Importantly, co-localization of prostate, stomach, duodenum, ileum, cecum, colon, pancreas, liver, TRPV5 and TRPV6 in the kidney may have significant func- spleen, brain, lung, heart, and skeletal muscle was sampled. Tissues Ϫ tional relevance, because it was recently shown that TRPV5 were subsequently stored at 80°C. The animal ethics board of the and TRPV6 can form heterotetrameric Ca2ϩ channels with University Medical Center Nijmegen approved all experimental procedures. distinct functionality (17). The aim of the present study was, therefore, to determine the localization and regulation of TRPV6 in the kidney. Immunohis- Immunohistochemistry tochemistry was performed using a new TRPV6 to The renal localization of TRPV6 was assessed by immunohisto- elucidate the tubular and subcellular localization of this epithelial chemistry, including co-localization studies using proteins with estab- Ca2ϩ channel. In animal experiments, the effect of 1,25(OH) D lished distribution patterns as markers for distinct nephron segments. 2 3 Immunohistochemical staining was performed as described previ- on renal TRPV6 mRNA and protein expression was studied using ously (18). In short, either co-immunohistochemical staining or stain- real-time quantitative polymerase chain reaction (PCR) analysis ing of serial sections for TRPV6 with TRPV5, calbindin-D28K, cal- and semiquantitative immunohistochemistry. ϩ Ϫ bindin-D9K, the -sensitive Na -Cl cotransporter (NCC), -2 (AQP2), and peanut lectin-binding intercalated cells was Materials and Methods performed on 7-␮m sections of fixated frozen kidney samples. Generation and Characterization of Affinity-Purified TRPV5 staining involved immersion of the kidney sections in boiled Rabbit TRPV6 Antibody citrate target retrieval buffer (0.01 M sodium citrate and 0.01 M citric Antiserum against TRPV6 was obtained by immunization of rab- acid, pH 6.0), which was then left to cool for 30 min, and subsequent bits with 400 ␮g of a keyhole limpet hemocyanin-coupled synthetic incubation in 0.3% (vol/vol) H2O2 in buffer (0.15 M NaCl, 0.1 M representing the last 15 amino acids of the C-terminal tail of Tris-HCl, pH 7.5) for 30 min. Sections were incubated for 16 h at 4°C with the primary : affinity-purified rabbit TRPV6 antibody mouse TRPV6 (NH2-INRGLEDGEGWEYQI-COOH). The anti- serum was affinity-purified according to standard procedures. The (1:25), affinity-purified guinea pig TRPV5 antibody (1:1000) (18), specificity of the affinity-purified rabbit TRPV6 antibody was as- guinea pig anti-calbindin-D9K (1:500) (12), mouse anti-calbindin- sessed by immunoblot analysis of Xenopus laevis oocytes heterolo- D28K (Swant, Bellinzona, Switzerland) (1:750), rabbit anti-NCC (1: gously expressing TRPV6 or TRPV5. In addition, the efficacy of the 200) (19), guinea pig anti-AQP2 (1:1000) (20), and FITC-coupled antibody to detect endogenous TRPV6 was assessed by immunoblot peanut lectin (1:100), respectively. In addition, immunohistochemis- analyses of mouse kidney and duodenum lysates. try of kidney sections using pre-immune serum and TRPV6 antiserum pre-absorbed for 1 h with the corresponding peptide was performed as negative controls. For detection of TRPV6, calbindin-D , calbindin- Animal Studies 9K D , NCC, and AQP2 sections were incubated with Alexa-conju- Experiment 1. 28K To study the nephron localization of TRPV6 by gated secondary antibodies. After incubation with biotin-coated goat immunohistochemistry, kidneys were isolated from C57BL6 mice anti–guinea pig secondary antibody, TRPV5 was visualized using a (kindly provided by Dr. J. Loffing, University of Zürich, Switzerland). tyramide signal amplification kit (NEN Life Science Products, The tissue fixation procedure consisted of anesthestizing the animals, Zaventem, Belgium). Images were made using a Nikon Diaphot after which the abdominal aorta was clamped downstream of the renal confocal laser scanning microscope (Tokyo, Japan; MRC-1000; Bio- arteries. Tubing was inserted at the level of the iliac bifurcation into Rad, Richmond, CA) and a Zeiss fluorescence microscope equipped the aorta, pushed up to the aortic clamp, and fixed by a ligature. The with a digital photo camera (Nikon DMX1200). For semiquantitative vena cava was opened, the aortic clamp was removed, and a fixative determination of protein levels, images were analyzed with the Image (50 ml) was allowed to flush the mouse vasculature under Pro Plus 4.1 image analysis software (Media Cybernetics, Silver high pressure. The fixative consisted of 3% paraformaldehyde (vol/ Spring, MD), resulting in quantification of the protein levels as the vol) and 0.05% picric acid (vol/vol) dissolved in a 3:2 mixture of 0.1 mean of integrated optical density (IOD). M cacodylate buffer (pH 7.4) and 10% hydroxyethyl starch (vol/vol) in saline (10). After 5-min fixation, the kidneys were rinsed at hy- drostatic pressure by perfusion for 5 min with the cacodylate buffer. Real-Time Quantitative PCR Analysis Kidneys were subsequently collected and stored at Ϫ80°C until fur- Total RNA was extracted from kidney and other tissues using ther processing. TriZol Total RNA Isolation Reagent (Life Technologies BRL, Breda,

Experiment 2. To study the effect of 1,25(OH)2D3 on renal The Netherlands). The obtained RNA was subjected to DNAse treat- TRPV6 mRNA and protein expression, C57BL6 mice were intraperi- ment to prevent genomic DNA contamination. Thereafter, 2 ␮gof toneally injected for 2 days with either 1,25(OH)2D3 at a dose of 100 RNA was reverse transcribed by Moloney-Murine Leukemia Virus- ng daily (n ϭ 6) or vehicle only (n ϭ 6). Thereafter, animals were Reverse Transcriptase (M-MLV-RT; Life Technologies BRL) as de- sacrificed and kidneys were sampled and either immediately frozen in scribed previously (18). The obtained cDNA was used to determine liquid nitrogen or fixated for immunohistochemistry by immersion in TRPV6 and TRPV5 mRNA expression levels as well as mRNA levels of 1% (wt/vol) periodate-lysine-paraformaldehyde (PLP) for2hand the housekeeping hypoxanthine-guanine phosphoribosyl transferase 15% (wt/vol) in phosphate-buffered saline overnight. Samples (HPRT) as an endogenous control. PCR primers and fluorescence probes were subsequently stored at Ϫ80°C until TRPV6 and TRPV5 mRNA were designed using the computer program Primer Express (Applied and protein expression levels were determined by real-time quantita- Biosystems, Foster City, CA) and purchased from Biolegio (Malden, The tive PCR and semiquantitative immunohistochemistry, respectively. Netherlands). The primer sequences for TRPV6 were 5'-ATCCGCCGC- Experiment 3. To evaluate the mRNA expression of TRPV6 TATGCACA-3', 5'-AGTTTTTCTCCTGA ATCTTTTTCCA-3', and 5'- relative to TRPV5 expression in kidney and to determine the quanti- TTCCAGCAACAAGATGGCCTCTACTCTGA-3' for the probe. tative expression in various other tissues, a cDNA panel was con- TRPV5 primer and probe sequences were 5'-CGTTGGTTCTTACGG J Am Soc Nephrol 14: 2731–2740, 2003 Epithelial Ca2ϩ Channel TRPV6 in Kidney 2733

GTTGAAC-3', 5'-GTTTGGAGAACCACAGAGCCTCTA-3', and 5'- TGTTTCTCAGATA GCTGCTCTTGTACTTCCTCTTTGT-3'. For HPRT, these primers were, respectively, 5'-TTATCAGACTGAA- GAGCTACTGTAATGATC-3', 5'-TTACCAGTGGTCAATTATTA CTTCAACAATC-3', and 5'-TGAGAGATCATCTCCACCAATA- ACTTTTATTGTCCC-3' for the probe. Expression levels were quanti- fied by real-time quantitative PCR on an ABI Prism 7700 Sequence Detection System (PE Biosystems, Rotkreuz, Switzerland).

TRPV6 Analysis

The general mechanism by which 1,25(OH)2D3 induces gene tran- scription involves direct interaction of the with regulatory domains in the promoter region of the gene known as vitamin D responsive elements (VDRE). To identify putative VDRE, the nucleotide sequence upstream of the ATG codon in the mouse TRPV6 (mTRPV6) gene was analyzed (Genbank accession no. NT039341).

Statistical Analyses Data are expressed as means Ϯ SEM. Statistical comparisons were tested by one-way ANOVA and Fisher multiple comparison. P values less than 5% were considered statistically significant. All analyses were performed using the Statview Statistical Package software (Pow- er PC version 4.51; Statview, Berkeley, CA) on a Macintosh computer.

Results Generation and Characterization of Affinity-Purified Rabbit TRPV6 Antibody Figure 1. Characterization of the affinity-purified rabbit TRPV6 an- Figure 1 shows immunoblots of lysates of Xenopus laevis tibody. Immunoblot of total membrane preparations of Xenopus laevis oocytes heterologously expressing TRPV6 or TRPV5 probed with oocytes heterologously expressing TRPV6 or TRPV5. The affinity-purified rabbit TRPV6 antibody (A) and affinity-purified affinity-purified TRPV6 antibody detected two specific bands guinea pig TRPV5 antibody (B). Non-injected, non-injected oocytes; of approximately 75 kD and 85 to 100 kD in size correspond- TRPV6, oocytes injected with 10 ng of TRPV6 cRNA; TRPV5, ing to the core and complex glycosylated proteins, respec- oocytes injected with 10 ng of TRPV5 cRNA. tively, as shown previously (17), whereas the affinity-purified TRPV5 antibody failed to detect these specific bands. Con- versely, in lysates of TRPV5 expressing Xenopus laevis oo- negative controls, including immunohistochemistry on kidney cytes, specific signals were not detected using the affinity- sections with either antiserum pre-absorbed for 1 h with 10 purified TRPV6 antibody. However, the affinity-purified ␮g/ml peptide (Figure 2E), pre-immune serum, or omission of TRPV5 antibody recognized two specific bands of 75 kD and the primary antibody (data not shown), were devoid of 85 to 100 kD, corresponding to the core and complex glyco- staining. sylated protein, respectively. Therefore, the applied antibodies Staining of serial kidney sections for TRPV5 and TRPV6 did not cross-react, indicating that both antibodies were chan- demonstrated that these epithelial Ca2ϩ channels show consid- nel-specific. In addition, immunoblot analysis of mouse kidney erable co-expression in renal cortex. All TRPV5-immunoposi- and duodenum lysates did not result in a specific immunopo- tive tubules expressed TRPV6, but a substantial number of sitive signal using either affinity-purified antibody, probably TRPV6-immunopositive tubules lacked TRPV5 immunoreac- due to the relatively low abundance of these channels (data not tivity (Figure 3A). In mice, TRPV5 expression is known to shown). start at the transition from DCT1 to DCT2, continuing into CNT (10,18). Calbindins are also expressed in DCT2 and the Immunohistochemical Localization of TRPV6 in Kidney principal cells of CNT, extending farther into CCD, and are To elucidate the renal segmental localization of TRPV6, involved in intracellular buffering and facilitated diffusion of immunohistochemical staining of mouse kidney sections with Ca2ϩ (10). Figure 3B shows that TRPV6 co-localizes with the TRPV6 antibody was performed. Distinct immunopositive calbindin-D28K, although a portion of TRPV6-positive tubules staining for TRPV6 was detected in kidney cortex (Figure 2, A did not express calbindin-D28K (data not shown). Co-staining and B) and, in addition, in renal medulla (Figure 2, C and D). for TRPV6 and calbindin-D9K also showed a considerable Importantly, in both cortex and medulla, TRPV6 immunola- co-expression of both proteins (Figure 4D). However, a num- beling was predominantly confined to the apical domain of the ber of TRPV6-immunopositive cells lacked calbindin-D9K.In tubular cells, suggesting apical membrane localization. All line with previous studies, renal medulla was devoid of any 2734 Journal of the American Society of Nephrology J Am Soc Nephrol 14: 2731–2740, 2003

Figure 2. Immunohistochemical localization of TRPV6 in mouse kidney. Overview of kidney cortex showing TRPV6 immunopositive staining (A). Higher magnification showing subcellular localization of TRPV6 (B). Overview of renal medulla showing TRPV6 immunopo- sitive staining (C). Subcellular localization of TRPV6 in the tubules of the renal medulla (D). Kidney cortex section stained with TRPV6 antiserum pre-absorbed for 1 h with 10 ␮g/ml peptide as a negative control (E). Figure 3. Immunohistochemical localization of TRPV6, TRPV5, cal- ϩ Ϫ bindin-D28K, and the Na -Cl cotransporter (NCC) in mouse kidney. Serial sections were stained with the affinity-purified rabbit TRPV6 TRPV5, calbindin-D9K, or calbindin-D28K immunopositive antibody (TRPV6) and affinity-purified guinea pig TRPV5 (TRPV5) signal (data not shown). antibody (A). Serial sections stained for TRPV6 and calbindin-D28K Serial kidney sections were stained for TRPV6 and the with rabbit anti-calbindin-D antiserum (B). Serial sections stained ϩ Ϫ 28K thiazide-sensitive Na -Cl cotransporter (NCC) that is exclu- for TRPV6 and NCC using rabbit anti-NCC antiserum (C and D). The sively expressed in DCT and functionally characterizes this asterisks depict tubule cells lacking co-expression for the respective segment (Figures 3, C and D) (9,10). In addition to tubules in serially stained proteins. which both proteins co-localized, NCC expression was not detected in a significant number of tubules that exhibited strong apical TRPV6 immunoreactivity. Furthermore, some To establish the identity of the additional TRPV6 positive NCC-positive tubules lacked TRPV6 immunostaining, sub- tubules, co-staining of mouse kidney sections for TRPV6 and stantiating that TRPV6 is not expressed in DCT1. Taken to- AQP2, characterizing the CNT and collecting ducts, was per- gether, these data demonstrate that TRPV6 expression starts formed (9,21). A bright immunopositive TRPV6 signal co- together with TRPV5 at the transition from DCT1 to DCT2. In localized with AQP2 in the cortical collecting ducts (CCD) contrast to TRPV5 expression, TRPV6 distribution extended (Figure 4A), outer medullary collecting ducts (OMCD), and into consecutive nephron segments. inner medullary collecting ducts (IMCD) (Figure 4B). Higher J Am Soc Nephrol 14: 2731–2740, 2003 Epithelial Ca2ϩ Channel TRPV6 in Kidney 2735

Figure 4. Immunohistochemical localization of TRPV6, aquaporin-2 (AQP2), and calbindin-D9K in mouse kidney. Double staining of affinity-purified rabbit TRPV6 antibody with rabbit anti-AQP2 antiserum (A, B, and C), rabbit anti-calbindin-D9K antiserum (D), and peanut lectin (E). The arrows in panels C and D mark AQP2 and calbindin-D9K negative cells, suggesting intercalated cells, substantiated by peanut lectin binding in panel E. magnification of these AQP2/TRPV6 co-expressing cells the apical side in some of these cells. Taken together, these showed that, although AQP2 is expressed strictly at the apical data indicate that TRPV6 distribution extends from DCT2 into membrane, TRPV6 shows additional intracellular staining at IMCD. In addition, Figures 4C and 4D show that TRPV6 was 2736 Journal of the American Society of Nephrology J Am Soc Nephrol 14: 2731–2740, 2003 detected along the apical domain in the intercalated cells of the CD and CNT, respectively (arrows). Co-immunohistochemical staining of TRPV6 with peanut lectin, identifying over 90% of the intercalated cell population, confirmed the strictly apical TRPV6 expression in this cell type (Figure 4F) (22). Therefore, in addition to the principal cells, TRPV6 was clearly expressed in the intercalated cells, which characteristically lack expres- sion of Ca2ϩ transport proteins.

Identification of Putative VDRE in the TRPV6 Promoter Region To identify putative VDRE, the mTRPV6 5'UTR was ana- lyzed for the presence of putative hexameric VDRE consensus sequence repeats, known to be involved in the transcriptional regulation of genes by 1,25(OH)2D3. VDRE have been re- ported to consist of two imperfect repeats separated by a limited number of nucleotide pairs known to contain the hex- americ RRKNSA (R ϭ AorG,Kϭ GorT,Sϭ CorG) consensus sequence (23,24). Figure 5 depicts four putative VDRE that were detected within 1500 bp upstream of the mTRPV6 ATG codon (black boxes).

Regulation of Renal TRPV6 Expression by 1,25(OH)2D3 The regulation of renal TRPV6 expression by 1,25(OH)2D3 was investigated in mice. TRPV6 mRNA copy numbers, as determined by real-time quantitative PCR analysis, were sig- nificantly increased in 1,25(OH)2D3-treated animals relative to controls (Figure 6A). This increased mRNA expression was accompanied by a significantly enhanced TRPV6 immunoflu-

Figure 6. Regulation of TRPV6 expression by 1,25(OH)2D3 in mouse kidney. TRPV6 mRNA expression was quantified by real-time quan- titative polymerase chain reaction (PCR) analysis (A). TRPV6 protein expression in kidney cortex (B) and medulla (C) as determined by immunohistochemistry. Representative immunohistochemical stain-

ing of kidney cortex showing the effect of 1,25(OH)2D3 on TRPV6 expression (D). TRPV5 mRNA (E) and protein (F) expression in kidney cortex. Protein expression is presented as integrated optical density (IOD). Data are presented as means Ϯ SEM. * P Ͻ 0.05 versus controls.

orescence in kidney cortex (Figure 6B) and medulla (Figure 6C), indicating that cortical and medullary TRPV6 expression

is positively regulated by 1,25(OH)2D3. Figure 6D shows representative immunohistochemical staining for TRPV6, il- Figure 5. Putative vitamin D–responsive elements (VDRE) in the lustrating the upregulation of TRPV6 protein expression by mouse TRPV6 promoter region. Nucleotide sequence upstream of the 1,25(OH)2D3 in kidney cortex. Furthermore, in concordance start codon of the mouse TRPV6 gene (Genbank accession no. with previously reported data, TRPV5 mRNA and proteins NT039341), which was probed for VDRE. Putative VDRE are indi- levels were significantly increased by 1,25(OH)2D3 (Figure 6, cated by black boxes. E and F). J Am Soc Nephrol 14: 2731–2740, 2003 Epithelial Ca2ϩ Channel TRPV6 in Kidney 2737

Quantification of TRPV6 mRNA Expression in Kidney denum, cecum Ͼ prostate, brain, liver Ͼ heart, pancreas, and Other Tissues spleen (Figure 7B). To study the expression of TRPV6 and TRPV5 in various mouse tissues, including kidney, a mouse cDNA panel was Discussion ϩ constructed. Subsequently, TRPV6 and TRPV5 mRNA levels The present study demonstrated that the epithelial Ca2 were quantified in this panel by real-time PCR analysis and channel TRPV6 is expressed along the apical domain of DCT2 normalized for cDNA input. Figure 7A demonstrates that through IMCD in mouse kidney. Together with the extensive TRPV6 mRNA was identified in various tissues and, in the co-localization with the proteins known to be involved in renal 2ϩ order of decreasing expression level, were detected: prostate Ͼ active Ca transport, the positive regulation of TRPV6 by 2ϩ stomach, brain Ͼ lung Ͼ duodenum, cecum, heart, kidney, 1,25(OH)2D3 suggests a role for TRPV6 in Ca reabsorption. bone Ͼ colon Ͼ skeletal muscle Ͼ pancreas. Likewise, However, TRPV6 was also clearly detected in nephron seg- TRPV5 was expressed in kidney Ͼ bone Ͼ stomach Ͼ duo- ments and extrarenal tissues not known to accommodate active Ca2ϩ absorption. Therefore, additional physiologic functions for TRPV6 are feasible. By detailed immunohistochemical analysis, the renal seg- mental and subcellular localization of TRPV6 was elucidated; the results are summarized in Figure 8. In both renal cortex and medulla, TRPV6 expression was predominantly confined to the apical membrane. TRPV6 co-localized with TRPV5 and the calbindins in DCT2 and CNT. It was previously shown that the Naϩ/Ca2ϩ‘exchanger (NCX1) and the plasma membrane Ca2ϩ-ATPase (PMCA), constituting the basolateral Ca2ϩ ex- trusion machinery, are also confined to these nephron segments (9,10). In addition, TRPV6 co-localized with AQP2 in the principal cells of CCD, OMCD, and IMCD. Interestingly, the intercalated cells in CNT and collecting duct also displayed apical TRPV6 expression. Thus, it was concluded that TRPV6 is consistently present along the luminal membranes of DCT2 through IMCD.

1,25(OH)2D3 is one of the most important regulators of Ca2ϩ homeostasis, and vitamin D receptors are primarily ex- pressed in kidney, small intestine, and bone (25–28). The

regulation of TRPV6 expression by 1,25(OH)2D3 was there- fore investigated in kidney. In mice exposed for 48 h to

Figure 8. Renal distribution of TRPV6 in mouse kidney. Summary of the renal distribution of TRPV6 as determined by immunohistochem- Figure 7. Quantification of tissue distribution of TRPV6 and TRPV5. istry. DCT, distal convoluted tubule; CNT, connecting tubule; CCD, TRPV6 (A) and TRPV5 (B) mRNA expression levels in a large set of cortical collecting duct; OMCD, outer medullary collecting duct; mouse tissues were determined by real-time quantitative PCR analy- IMCD, inner medullary collecting duct; PMCA, plasma membrane sis. Copy numbers were normalized for the cDNA concentration of Ca2ϩ-ATPase; NCX, Naϩ/Ca2ϩ exchanger; NCC, Naϩ-ClϪ cotrans- the input. Data are presented as means Ϯ SEM. porter; AQP2, aquaporin-2. 2738 Journal of the American Society of Nephrology J Am Soc Nephrol 14: 2731–2740, 2003

1,25(OH)2D3, a modest but significant upregulation of renal collecting ducts, which are responsible for the fine-tuning of TRPV6 mRNA expression was observed, accompanied by a renal Naϩ excretion and are highly permeable to in the marked increase in tubular TRPV6 protein abundance in both presence of , displayed distinct apical TRPV6 ex- kidney cortex and medulla. These data demonstrated that pression. Second, intercalated cells, known to be involved in

1,25(OH)2D3 stimulates renal TRPV6 transcription and sug- urinary acidification, also expressed TRPV6. The absence of 2ϩ gested that additional translational regulation by 1,25(OH)2D3 supportive Ca transport proteins along the larger part of the occurs. In general, the genomic mechanism of 1,25(OH)2D3 CD, as confirmed in this study by the lack of immunopositive action has been shown to involve direct interaction of the staining for calbindins in OMCD and IMCD, precludes in- activated vitamin D receptor complex with regulatory domains volvement of TRPV6 in transcellular Ca2ϩ reabsorption in in the promoter region of genes, known as VDRE (23,24). these tubules (10). The consistent predominant apical localiza- Promoter sequence analysis described in the present study tion would, however, imply that TRPV6 has a functional role identified several putative VDRE upstream of the mTRPV6 as apical Ca2ϩ entry channel. Hypothetically, TRPV6-medi- start codon. The promoter sequences of the genes encoding the ated Ca2ϩ influx could affect transport processes in these cells, vitamin D–responsive Ca2ϩ transport proteins TRPV5, calbi- for instance as part of a hormonal signaling cascade. Vasopres- 2ϩ ndin-D9K and calbindin-D28K, were previously shown to con- sin was shown to induce Ca influx across the apical mem- tain several VDRE, and stimulation of the expression of these brane of renal collecting duct cells, and alterations in intracel- 2ϩ ϩ transporters in the kidney by 1,25(OH)2D3 has been repeatedly lular Ca levels are known to affect Na reabsorption in these confirmed (11,12,29,30). tubules (34–36). In addition, it has been suggested that the The predominantly apical localization of TRPV6, which has action of other hormones, including atrial natriuretic peptide previously been shown to constitute a highly Ca2ϩ-selective (ANP), parathyroid hormone (PTH), , and prosta- channel, is in line with a role as apical Ca2ϩ entry mech- glandins on transport processes involve Ca2ϩ signaling (37– anism in kidney (8). The co-localization of TRPV6 with the 39). The tonicity of the pre-urine osmotically challenges tubu- other key players in active Ca2ϩ transport in DCT2 and CNT, lar epithelial cells constantly, and the resulting cell volume the main sites of active Ca2ϩ reabsorption, in conjunction with regulation is facilitated by Ca2ϩ signaling, which has been 2ϩ the positive regulation by 1,25(OH)2D3 in kidney clearly sup- shown to encompass Ca release from intracellular stores ports our hypothesis that TRPV6 could have functional impor- followed by Ca2ϩ influx from the extracellular compartment 2ϩ 2ϩ tance in 1,25(OH)2D3–regulated transcellular Ca reabsorp- (40,41). The identity of the apical Ca entry mechanism in the tion (10). TRPV6 was originally cloned from duodenum and aforementioned processes has not been elucidated. Therefore, was recently localized to the brush border membrane of the TRPV6 might be a potential candidate. small intestine, in which calbindin-D9K and PMCA are also In this study, TRPV6 and TRPV5 expression levels were known to be expressed (3,14,18). The regulation of TRPV6 quantified in a large set of tissues. Our results showed that 2ϩ expression by dietary Ca and 1,25(OH)2D3 in duodenum and TRPV6 mRNA is expressed in nearly all tissues studied but intestinal cell lines has further emphasized the involvement of that its expression levels vary greatly. A robust TRPV6 ex- this channel in dietary Ca2ϩ absorption (15,30–33). Previ- pression was detected in prostate. Although the exact function ously, an essential role was not attributed to TRPV6 in kidney, in this organ remains to be elucidated, previous reports have while TRPV5 was generally considered as the rate-limiting suggested that TRPV6 expression correlates with prostate car- Ca2ϩ entry step in renal transcellular Ca2ϩ reabsorption. It is, cinoma tumor grade (42,43). In addition, TRPV6 was ex- however, tempting to speculate on the basis of the present renal pressed at lower levels in both kidney and duodenum. The localization data that TRPV6 could facilitate active Ca2ϩ (re)- functional relevance of TRPV6 in dietary Ca2ϩ absorption has absorption in both kidney and intestine. been previously illustrated. Therefore, the detection of compa- We recently demonstrated that the epithelial Ca2ϩ channels rable expression levels in kidney further substantiated that are present in the plasma membrane as functional tetrameric renal TRPV6 expression is indeed quantitatively significant. complexes (17). Moreover, TRPV5/TRPV6 heterotetramers Of note, substantial TRPV6 expression was detected in bone, were shown to form functional Ca2ϩ channels in mammalian an important reservoir of rapidly exchangeable Ca2ϩ, suggest- and Xenopus laevis oocyte expression systems. Depending on ing a role in bone mineralization. Interestingly, TRPV6 mRNA their exact subunit composition, these heterotetramers exhib- was also abundantly expressed in stomach, which is not known ited distinct electrophysiologic characteristics. The co-localiza- as a site of Ca2ϩ absorption. This might point to an additional tion of TRPV6 and TRPV5 along the apical membrane of role for TRPV6 in the , distinct from DCT2 and CNT, which was demonstrated in this study, sug- transcellular Ca2ϩ absorption. Furthermore, TRPV6 was de- gests that these heterotetramers can potentially be formed in tected at significant levels in brain, lung, and heart, but the the kidney. Hypothetically, co-expression of both epithelial functional relevance in these tissues is not clearly envisaged. Ca2ϩ channels may provide these Ca2ϩ transporting tubules TRPV5 expression levels were remarkably high in kidney, but with a pleiotropic set of heterotetrameric channels to cover a TRPV5 was detected in various other tissues as well, including broad range of Ca2ϩ transport kinetics. bone and intestine, albeit at considerably lower levels. Given Interestingly, TRPV6 expression was also detected in the recent finding that both epithelial Ca2ϩ channels can com- nephron segments not generally implicated in active Ca2ϩ bine into heterotetramers with novel functional properties, the reabsorption. First, principal cells in cortical and medullary variation in relative expression of TRPV6 and TRPV5 in the J Am Soc Nephrol 14: 2731–2740, 2003 Epithelial Ca2ϩ Channel TRPV6 in Kidney 2739 tissues studied might reflect differential regulation to fine-tune 9. Reilly RF, Ellison DH: Mammalian distal tubule: Physiology, Ca2ϩ channel kinetics (17). Altogether, these data indicate that, pathophysiology, and molecular anatomy. Physiol Rev 80: 277– in addition to facilitating active Ca2ϩ transport in duodenum 313, 2000 and kidney, the epithelial Ca2ϩ channels might also have 10. Loffing J, Loffing-Cueni D, Valderrabano V, Klausli L, Hebert additional functional roles in the body. SC, Rossier BC, Hoenderop JG, Bindels RJ, Kaissling B: Orga- In conclusion, the localization of TRPV6 to the apical do- nization of the mouse distal nephron: distributions of transcellu- lar and sodium transport pathways. Am J Physiol Renal main of cells lining DCT2 to IMCD and the positive regulation Physiol 281: F1021–F1027, 2001 by 1,25(OH) D substantiate the functional relevance of 2 3 11. Hoenderop JG, Muller D, Van Der Kemp AW, Hartog A, Suzuki TRPV6 in kidney. The co-expression of TRPV6 and TRPV5 M, Ishibashi K, Imai M, Sweep F, Willems PH, Van Os CH, et 2ϩ implies that a pleiotropic set of heterotetrameric Ca channels al: controls the epithelial in kidney. 2ϩ could facilitate renal active Ca reabsorption. However, the J Am Soc Nephrol 12: 1342–1349, 2001 precise role of TRPV6 in kidney remains to be established, 12. Hoenderop JG, Dardenne O, van Abel M, van der Kemp AW, ϩ particularly because this epithelial Ca2 channel was also Van Os CH, St-Arnaud R, Bindels RJ: Modulation of renal Ca2ϩ ϩ detected in nephron segments not involved in Ca2 reabsorp- transport protein genes by dietary Ca2ϩ and 1,25-dihydroxyvi- ␣ tion. Given the widespread distribution of TRPV6 throughout tamin D3 in 25-hydroxyvitamin D3-1 -hydroxylase knockout the body, additional functions for TRPV6 are indeed conceiv- mice. Faseb J 16: 1398–1406, 2002 able, including a role in Ca2ϩ signaling. 13. van Abel M, Hoenderop JG, Dardenne O, St-Arnaud R, Van Os C, van Leeuwen JP, Bindels RJ: 1,25(OH)2: D3-independent stimulatory effect of on the expression of ECaC1 in Acknowledgments kidney. J Am Soc Nephrol 13: 2102–2109, 2002 14. Zhuang L, Peng JB, Tou L, Takanaga H, Adam RM, Hediger The authors thank Dr. J. Loffing, University of Zürich, Switzer- MA, Freeman MR: Calcium-selective , CaT1, is land, for kindly providing sections of perfused mouse kidneys. This apically localized in gastrointestinal tract epithelia and is aber- project was financially supported by the Dutch Kidney Foundation rantly expressed in human malignancies. Lab Invest 82:1755– (C00.1881) and the Dutch Organization for Scientific Research (Zon- 1764, 2002 Mw 016.006.001). 15. van Abel M, Hoenderop JG, van der Kemp AW, van Leeuwen JP, Bindels RJ: Regulation of the epithelial Ca2ϩ channels in small intestine as studied by quantitative mRNA detection. Am J References Physiol Gastrointest Liver Physiol 285: G78–G85, 2003 1. Hoenderop JG, Nilius B, Bindels RJ: Molecular mechanisms of 16. Peng JB, Brown EM, Hediger MA: Structural conservation of the active Ca2ϩ reabsorption in the distal nephron. Ann Rev Physiol genes encoding CaT1. CaT 2 and related cation channels. 64: 529–549, 2002 Genomics 76: 99–109, 2001 2. Hoenderop JG, van der Kemp AW, Hartog A, van de Graaf SF, 17. Hoenderop JG, Voets T, Hoefs S, Weidema AF, Prenen J, Nilius van Os CH, Willems PH, Bindels RJ: Molecular identification of 2ϩ B, Bindels RJ: Homo- and heterotetrameric architecture of the the apical Ca channel in 1,25-dihydroxyvitamin D -responsive ϩ 3 epithelial Ca2 channels, TRPV5 and TRPV6. Embo J 22: 776– epithelia. J Biol Chem 274: 8375–8378, 1999 785, 2003 3. Peng JB, Chen XZ, Berger UV, Vassilev PM, Tsukaguchi H, 18. Hoenderop JG, Hartog A, Stuiver M, Doucet A, Willems PH, Brown EM, Hediger MA: Molecular cloning and characteriza- Bindels RJ: Localization of the epithelial Ca2ϩ channel in rabbit tion of a channel-like transporter mediating intestinal calcium kidney and intestine. J Am Soc Nephrol 11: 1171–1178, 2000 absorption. J Biol Chem 274: 22739–22746, 1999 4. Muller D, Hoenderop JG, Meij IC, van den Heuvel LP, Knoers 19. Nijenhuis T, Hoenderop JG, Loffing J, van der Kemp AW, Van Os C, Bindels RJ: Thiazide-induced hypocalciuria is accompa- NV, den Hollander AI, Eggert P, Garcia-Nieto V, Claverie- 2ϩ Martin F, Bindels RJ: Molecular cloning, tissue distribution, and nied by a decreased expression of Ca transporting proteins in chromosomal mapping of the human epithelial Ca2ϩ channel the distal tubule. Kidney Int 64: 555–564, 2003 (ECAC1). Genomics 67: 48–53, 2000 20. Mulders SM, Bichet DG, Rijss JP, Kamsteeg EJ, Arthus MF, 5. Muller D, Hoenderop JG, Merkx GF, van Os CH, Bindels RJ: Lonergan M, Fujiwara M, Morgan K, Leijendekker R, van der Gene structure and chromosomal mapping of human epithelial Sluijs P, et al: An aquaporin-2 water channel mutant which calcium channel. Biochem Biophys Res Commun 275: 47–52, causes autosomal dominant nephrogenic is 2000 retained in the Golgi complex. J Clin Invest 102: 57–66, 1998 6. Peng JB, Chen XZ, Berger UV, Weremowicz S, Morton CC, 21. Coleman RA, Wu DC, Liu J, Wade JB: Expression of Vassilev PM, Brown EM, Hediger MA: Human calcium trans- in the renal connecting tubule. Am J Physiol Renal Physiol 279: port protein CaT1. Biochem Biophys Res Commun 278: 326– F874–F883, 2000 332, 2000 22. Emmons C, Kurtz I: Functional characterization of three inter- 7. Hoenderop JG, Vennekens R, Muller D, Prenen J, Droogmans G, calated cell subtypes in the rabbit outer cortical collecting duct. Bindels RJ, Nilius B: Function and expression of the epithelial J Clin Invest 93: 417–423, 1994 Ca2ϩ channel family: comparison of the mammalian epithelial 23. Carlberg C: Mechanisms of nuclear signaling by vitamin D3. Ca2ϩ channel 1 and 2. J Physiol 537: 747–761, 2001 Interplay with retinoid and hormone signalling. Eur 8. Vennekens R, Hoenderop JG, Prenen J, Stuiver M, Willems PH, J Biochem 231: 517–527, 1995 Droogmans G, Nilius B, Bindels RJ: Permeation and gating 24. Christakos S, Raval-Pandya M, Wernyj RP, Yang W: Genomic properties of the novel epithelial Ca2ϩ channel. J Biol Chem 275: mechanisms involved in the pleiotropic actions of 1,25-dihy- 3963–3969, 2000 droxyvitamin D3. Biochem J 316: 361–371, 1996 2740 Journal of the American Society of Nephrology J Am Soc Nephrol 14: 2731–2740, 2003

25. Reichel H, Koeffler HP, Norman AW: The role of the vitamin D collecting system: effects on cAMP and cytosolic Ca2ϩ. Pflugers endocrine system in health and disease. N Engl J Med 320: Arch 433: 109–115, 1996 980–991, 1989 35. Frindt G, Windhager EE: Ca2ϩ-dependent inhibition of sodium 26. Friedman PA, Gesek FA: Cellular calcium transport in renal transport in rabbit cortical collecting tubules. Am J Physiol 258: epithelia: measurement, mechanisms, and regulation. Physiol F568–F582, 1990 Rev 75: 429–471, 1995 36. Star RA, Nonoguchi H, Balaban R, Knepper MA: Calcium and 27. Brown AJ, Dusso A, Slatopolsky E: Vitamin D. Am J Physiol cyclic adenosine monophosphate as second messengers for va- 277: F157–F175, 1999 sopressin in the rat inner medullary collecting duct. J Clin Invest 28. Kumar R, Schaefer J, Grande JP, Roche PC: Immunolocalization 81: 1879–1888, 1988 of calcitriol receptor, 24-hydroxylase cytochrome P-450, and 37. Oliveira-Souza M, Mello-Aires M: Effect of arginine vasopressin 2ϩ calbindin D28k in human kidney. Am J Physiol 266: F477–F485, and ANP on intracellular pH and cytosolic free [Ca ] regulation 1994 in MDCK cells. Kidney Int 60: 1800–1808, 2001 29. Gross M, Kumar R: Physiology and biochemistry of vitamin 38. Hoenderop JG, Vaandrager AB, Dijkink L, Smolenski A, Gam- D-dependent calcium binding proteins. Am J Physiol 259: F195– baryan S, Lohmann SM, de Jonge HR, Willems PH, Bindels RJ: F209, 1990 Atrial natriuretic peptide-stimulated Ca2ϩ reabsorption in rabbit 30. Weber K, Erben RG, Rump A, Adamski J: Gene structure and kidney requires membrane-targeted, cGMP-dependent protein regulation of the murine epithelial calcium channels ECaC1 and kinase type II. Proc Natl Acad Sci USA 96: 6084–6089, 1999 2. Biochem Biophys Res Commun 289: 1287–1294, 2001 39. Ankorina-Stark I, Haxelmans S, Schlatter E: Receptors for bra- 2ϩ 31. Wood RJ, Tchack L, Taparia S: 1,25-Dihydroxyvitamin D3: dykinin and prostaglandin E2 coupled to Ca signalling in rat Increases the expression of the CaT1 epithelial calcium channel cortical collecting duct. Cell Calcium 22: 269–275, 1997 in the Caco-2 human intestinal cell line. BMC Physiol 1: 11, 40. Tinel H, Kinne-Saffran E, Kinne RK: Calcium signalling during 2001 RVD of kidney cells. Cell Physiol Biochem 10: 297–302, 2000 32. Fleet JC, Bradley J, Reddy GS, Ray R, Wood RJ: 1 alpha,25- 41. Woll E, Ritter M, Haller T, Volkl H, Lang F: Calcium entry

(OH)2-vitamin D3 analogs with minimal in vivo calcemic activ- stimulated by swelling of Madin-Darby Canine Kidney cells. ity can stimulate significant transepithelial calcium transport and Nephron 74: 150–157, 1996 mRNA expression in vitro. Arch Biochem Biophys 329: 228– 42. Wissenbach U, Niemeyer BA, Fixemer T, Schneidewind A, 234, 1996 Trost C, Cavalie A, Reus K, Meese E, Bonkhoff H, Flockerzi V: 33. van Cromphaut S, Dewerchin M, Hoenderop JG, Stockmans I, Expression of CaT-like, a novel calcium-selective channel, cor- van Herck E, Kato S, Bindels RJ, Collen D, Carmeliet P, Bouil- relates with the malignancy of prostate . J Biol Chem 276: lon R, et al: Active duodenal calcium absorption in vitamin D 19461–19468, 2001 receptor-knock out mice: Functional and molecular aspects. Proc 43. Peng JB, Zhuang L, Berger UV, Adam RM, Williams BJ, Brown Natl Acad Sci USA 98: 13324–13329, 2001 EM, Hediger MA, Freeman MR: CaT1 expression correlates 34. van Baal J, Raber G, de Slegte J, Pieters R, Bindels RJ, Willems with tumor grade in . Biochem Biophys Res Com- PH: Vasopressin-stimulated Ca2ϩ reabsorption in rabbit cortical mun 282: 729–734, 2001