Excision of Trpv6 Gene Leads to Severe Defects in Epididymal Ca2+ Absorption and Male Fertility Much Like Single D541A Pore Mutation*

Background: The TRPV6D541A pore mutation abrogates epididymal Ca2+ absorption causing hypofertility in mice, raising the possibility of residual TRPV6D541A channel activity. Results: Trpv6 deletion reduces fertility parameters to the same extent as the D541A pore mutation. Conclusion: The D541A pore mutation leads to complete inactivation of TRPV6 channels in epididymal epithelium. Significance: Targeted mutations in mice help to understand the function of TRPV6 proteins in native systems. Replacement of aspartate residue 541 by alanine (D541A) in the pore of TRPV6 channels in mice disrupts Ca2+ absorption by the epididymal epithelium, resulting in abnormally high Ca2+ concentrations in epididymal luminal fluid and in a dramatic but incomplete loss of sperm motility and fertilization capacity, raising the possibility of residual activity of channels formed by TRPV6D541A proteins (Weissgerber, P., Kriebs, U., Tsvilovskyy, V., Olausson, J., Kretz, O., Stoerger, C., Vennekens, R., Wissenbach, U., Middendorff, R., Flockerzi, V., and Freichel, M. (2011) Sci. Signal. 4, ra27). It is known from other cation channels that introducing pore mutations even if they largely affect their conductivity and permeability can evoke considerably different phenotypes compared with the deletion of the corresponding protein. Therefore, we generated TRPV6-deficient mice (Trpv6−/−) by deleting exons encoding transmembrane domains with the pore-forming region and the complete cytosolic C terminus harboring binding sites for TRPV6-associated proteins that regulate its activity and plasma membrane anchoring. Using this strategy, we aimed to determine whether the TRPV6D541A pore mutant still contributes to residual channel activity and/or channel-independent functions in vivo. Trpv6−/− males reveal severe defects in fertility and motility and viability of sperm and a significant increase in epididymal luminal Ca2+ concentration that is mirrored by a lack of Ca2+ uptake by the epididymal epithelium. Therewith, Trpv6 excision affects epididymal Ca2+ handling and male fertility to the same extent as the introduction of the D541A pore mutation, arguing against residual functions of the TRPV6D541A pore mutant in epididymal epithelial cells.

. It is known from other cation channels that introducing pore mutations even if they largely affect their conductivity and permeability can evoke considerably different phenotypes compared with the deletion of the corresponding protein. Therefore, we generated TRPV6-deficient mice (Trpv6 ؊/؊ ) by deleting exons encoding transmembrane domains with the pore-forming region and the complete cytosolic C terminus harboring binding sites for TRPV6-associated proteins that regulate its activity and plasma membrane anchoring. Using this strategy, we aimed to determine whether the TRPV6 D541A pore mutant still contributes to residual channel activity and/or channel-independent functions in vivo. Trpv6 ؊/؊ males reveal severe defects in fertility and motility and viability of sperm and a significant increase in epididymal luminal Ca 2؉ concentration that is mirrored by a lack of Ca 2؉ uptake by the epididymal epithelium. Therewith, Trpv6 excision affects epididymal Ca 2؉ handling and male fertility to the same extent as the introduction of the D541A pore mutation, arguing against residual functions of the TRPV6 D541A pore mutant in epididymal epithelial cells.
The maintenance of body Ca 2ϩ homeostasis is essential for many vital functions including neuronal excitability, muscle contraction, and bone formation. Ca 2ϩ acquisition in the body occurs via trans-and paracellular transport processes across the continuous layer of epithelial cells. About 10 years ago, transcripts of the structurally closely related TRPV6 and TRPV5 were identified in the epithelia of the kidney (2), in the duodenum (3), and in placenta, pancreatic acinar cells, and other exocrine glands (4 -7), and expression of the Trpv6 and Trpv5 cDNAs in HEK293 cells or other expression systems leads to the formation of cation channels with a high selectivity for Ca 2ϩ (3)(4)(5)8). These channels exhibit many features possessed by Ca 2ϩ transporters in epithelial cells (6,7): they mediate passive transport of Ca 2ϩ down the electrochemical gradient without energy consumption, and they are constitutively active (4,5,8). Accordingly, they were assumed to be epithelial Ca 2ϩ uptake channels.
Crucial for Ca 2ϩ permeation through TRPV6 and TRPV5 channels is a single aspartate residue within the pore-forming loop of both proteins; replacing this aspartate residue at position 541 in mouse TRPV6 or at position 542 in rabbit TRPV5 by an alanine residue renders the channels impermeable to Ca 2ϩ (9). We studied the impact of this mutation in vivo with a mouse model in which the D541A mutation of TRPV6 was introduced in the germ line by a gene targeting approach (1). Males homozygous for this single amino acid substitution (TRPV6 D541A/D541A ) exhibited a severely impaired fertility and a large reduction of motility and fertilization capacity of sperm despite intact spermatogenesis. An increase in Ca 2ϩ in spermatozoa is an important signal to promote their * This work was supported by the Deutsche Forschungsgemeinschaft (to M. F., V. F., P. W., and R. M.), Fonds der Chemischen Industrie and Sander-Stiftung (to V. F.), LOEWE program (to R. M.), and Landes-Offensive zur Entwicklung Wissenschaftlich-ö konomischer Exzellenz (LOEWE), Homburger Forschungsfö rderungsprogramm (HOMFOR) and Forschungsausschuss der Universität des Saarlandes (to M. F., V. F., P. W., and V. T.). □ S This article contains supplemental Movies S1 and S2. 1 These authors contributed equally to this work. 2  motility, capacitation, and the acrosome reaction (10,11), but Trpv6 transcripts were not detectable in spermatozoa or in the germinal epithelium, and there was no evidence for a cell-autonomous impairment of sperm Ca 2ϩ signaling (1). However, we identified Trpv6 transcripts in the epididymal epithelium and TRPV6 proteins in the apical membrane of this epithelium and showed that the luminal Ca 2ϩ concentration is increased by 10-fold in the caudal epididymal fluid of Trpv6 D541A/D541A males compared with that of wild-type mice. Additional measurements of Ca 2ϩ uptake from the epididymal fluid from Trpv6 D541A/D541A mice into the epididymis revealed a reduction of uptake by 7-8-fold compared with fluid from wild-type animals (1). Apparently, the intact TRPV6 proteins are essential components of Ca 2ϩ uptake channels in the epididymal epithelium that are responsible for the decrease of the Ca 2ϩ concentration in the epididymal fluid along the epididymal duct, generating a luminal Ca 2ϩ gradient with higher Ca 2ϩ concentration in the caput portion (proximal segments) of the epididymal duct and lower Ca 2ϩ concentrations in its caudal (distal) segments. It is well known that the composition of the epididymal fluid differs considerably among separate epididymal segments. This is caused by differences in secretory and absorptive activities of the epididymal epithelium (12)(13)(14)(15)(16), resulting in changes of the net water; the HCO 3 Ϫ reabsorption; the secretion of proteins; concentrations of Na ϩ , Cl Ϫ , and K ϩ ; and luminal acidification. Notably, the Ca 2ϩ concentration decreases markedly in the epididymal fluid along the epididymal duct toward its distal segments, the cauda epididymis (14). It is this luminal Ca 2ϩ gradient that is severely compromised by the TRPV6 D541A mutation, resulting in a dramatic but incomplete loss of sperm motility and fertilization capacity.
The direct electrophysiological recording of channel activity was not possible either from wild-type or from D541A mutant epididymal cells. Our own results and a survey of the literature on TRPV6 and its closest relative, TRPV5, revealed that direct recordings of TRPV6 and TRPV5 currents from acutely prepared primary cells expressing Trpv6 or Trpv5 transcripts have not been described. Instead, Ca 2ϩ uptake measurements similar to those we performed (17) have been used to identify (10,18) and characterize these channels in primary cells. Although Ca 2ϩ uptake was reduced by 7-8-fold in Trpv6 D541A/D541A mice, we could not exclude residual channel activity of the properly expressed and trafficked, yet mutated TRPV6 D541A proteins, which might in addition serve channel-independent functions as structural and/or scaffolding components (19) of the epithelial plasma membrane. Thus, the TRPV6 D541A mutation might not be sufficient to identify all TRPV6-related functions, and it is conceivable that a complete loss of TRPV6 proteins leads to additional phenotypes. Similarly, the GluR␦2 Q618R pore mutation, although rendering GluR␦2 channels nominally impermeable to Ca 2ϩ (20), did not cause the defects in synaptic plasticity of hippocampal neurons and motor coordination that were evoked by the complete inactivation of the GluR␦2 gene in mice (21). We therefore generated a second mouse line using a Cre-loxP-based gene targeting strategy to delete exons 13, 14, and 15 of the Trpv6 gene that encode part of the fifth and sixth transmembrane-spanning domains, the pore linker in between, and the complete cytosolic C terminus. By deleting almost one-third of the protein-encoding DNA, we not only eliminated an essential part of the ion-conducting pore but also disrupted essentially the predicted structure of the protein. We now show that mice homozygous for this gene excision were viable, and our analysis reveals that the Ca 2ϩ concentration in the caudal epididymal fluid and the motility, fertilization capacity, and viability of Trpv6 Ϫ/Ϫ sperm are affected to the same extent as in Trpv6 D541A/D541A mice. These results indicate that the TRPV6 deletion occurring in Trpv6 Ϫ/Ϫ mice does not further aggravate the phenotype observed in Trpv6 D541A/D541A mice, arguing against residual channel activity and against channel-independent scaffolding functions of the TRPV6 D541A protein that may influence epididymal Ca 2ϩ absorption or fertilization capacity indirectly.

EXPERIMENTAL PROCEDURES
All animal experiments were performed in accordance with German legislation on the protection of animals and were approved by the local ethics committee. The generation of TRPV6 Ϫ/Ϫ mice and the maintenance of the mouse line as well as the experiments performed were approved (reference number K110/180-07; approved on December 10, 2002, December 28, 2004, andJanuary 17, 2007) by the "Kreispolizeibehörde des Saarpfalz-Kreises, Deutschland" and was performed by M. F., V. F., P. W., U. K., J. O., S. B., K. F., and C. M. All these people got permission to perform the experiments described by the local ethics committee and "Kreispolizeibehörde des Saarpfalz-Kreises, Deutschland." They all have a long-standing experience and authorization in the generation and phenotypic analysis of transgenic mice (e.g. Ref. 1 and references therein).
Generation of TRPV6 Knock-out Mice-The 5Ј and 3Ј homology arms of the targeting vector were amplified from genomic DNA of R1 ES cells using Pfu polymerase. The 5Ј homology arm was cloned 5Ј of a loxP site followed by the genomic sequence containing exons 13, 14, and 15; an FRT 3 sequence-flanked PGK promotor-driven neomycin resistance gene cassette (neo), and a second loxP site. The sequence of the loxP site was inserted in the 16th intron of the Ephb6 gene, which is oriented tail to tail with Trpv6 in the mouse genome separated by 120 bp only. Accordingly, Cre-mediated excision of exons 13, 14, and 15 of the Trpv6 gene leads to deletion of exons 17 and 18 of the adjacent Ephb6 gene. The herpes simplex virus thymidine kinase cassette (HSVtk) and an eGFP cassette were introduced for negative selection. Gene targeting in R1 ES cells was performed as described (1). One of 327 double resistant colonies showed homologous recombination at the Trpv6 locus as confirmed by Southern blot hybridization with a 5Ј and 3Ј probe external to the targeting vector and a neo probe. Germ line chimeras were obtained by injection of correctly targeted ES cell clone 6F11 into C57Bl/6 blastocysts. Trpv6 Ϫ/Ϫ mice on the mixed (129/SvJ ϫ C57Bl/6N) background were compared with F1 offspring of 129SvJ ϫ C57Bl/6N matings unless stated otherwise. Routinely, mice were genotyped using PCR. Additionally, an independent Trpv6 Ϫ/Ϫ mouse line was generated from Trpv6 ϩ/D541A mice (ES cell clone 8E11b) by Cre-mediated excision of genomic sequences containing exons 13-15 and analyzed via computer-assisted sperm analysis. All animal experiments were performed according to the Guide for the Care and Use of Laboratory Animals published by the United States National Institutes of Health and were approved by the local ethics committee.

D541A Pore Mutation Inactivates TRPV6 Like Its Deletion
Expression Analysis of Trpv6-encoding Transcripts-Poly-(A) ϩ RNA (10 g) from epididymis and testis of wild-type and Trpv6 Ϫ/Ϫ mice was used for Northern blot analysis as described (17). Blots were hybridized with a randomly labeled cDNA probe comprising exons 13, 14, and 15 of the mTrpv6 cDNA (nucleotides 1741-2849; GenBank TM accession number AJ542487), and filters were exposed to x-ray films for 19 h. To analyze expression of Trpv6 transcripts in prostate, we prepared RNA from prostate using the RNeasy Mini kit (Qiagen) and performed one-step reverse transcription-PCR (RT-PCR; Invitrogen) using 20 ng of total RNA/reaction. The following intron-spanning primers were used: for amplification of a specific Trpv6 fragment comprising 258 bp: UK_V6_16 (5Ј-GTC TGG CAT CAG CCT CAG C-3Ј; exon 1) and UK_V6_33 (5Ј-CTC ACA TCC TTC AAA CTT GAG C-3Ј; exon 2); for amplification of full-length Trpv6 cDNA comprising 2312 bp: UK_V6_17 (5Ј-CAG GGT CGA GCC CAG TTG G-3Ј; located in the 5Ј-untranslated region of exon 1) and UK_V6_14 (5Ј-CTC GCA GGA TGA CCT TAG CTG-3Ј; located in the 3Ј-untranslated region of exon 15). RNA isolated from epididymis was used as a positive control.
Mating, Weight Gain, and Fertility Analysis-Fertility analysis was done basically as described in Weissgerber et al. (1). In brief, adult male Trpv6 ϩ/Ϫ and Trpv6 Ϫ/Ϫ mice were continuously housed with adult Trpv6 ϩ/Ϫ and Trpv6 Ϫ/Ϫ females over a period of 16 weeks, and the number and size of litters as well as the genotype of offspring were recorded. Considering 3 weeks per pregnancy and assuming that the adult female mouse gets pregnant within the 1st week after the birth of the litter, four litters can be expected per mating under optimal conditions during this time period of 16 weeks. In reality, we observed that each wild-type/wild-type mating produced slightly fewer numbers of litters, that is 3.375 litters in 16 weeks (27 litters per eight matings; see Ref. 1). All mating analyses described in this study and by Weissgerber et al. (1) were performed in our animal facility under identical conditions. For copulatory behavior analysis, the ratio of plug-positive females per mating was evaluated. After flushing of the oviduct of plug-positive females, collected embryos were distinguished from nonfertilized eggs by Hoechst 33258 staining of DNA. Body weight was analyzed weekly in wild-type, Trpv6 ϩ/Ϫ , and Trpv6 Ϫ/Ϫ littermates from Trpv6 ϩ/Ϫ intercrosses until the age of 21 weeks. To compare body weight development between Trpv6 Ϫ/Ϫ and Trpv6 D541A/D541A mice, body weight at the age of 4, 8, 12, and 16 weeks was divided by the body weight of the same mouse at the age of 1 week and normalized to the corresponding body weight ratio obtained from wild-type littermates. Isolation of spermatozoa, computer-assisted sperm analysis, in vitro fertilization, and sperm viability analysis were performed as described in detail (1).
Analysis of Segregation of the Trpv6 Ϫ Allele-To analyze whether the Trpv6 Ϫ allele was inherited in a mendelian ratio, Trpv6 ϩ/Ϫ males and females at the age of 3-4 months were intercrossed, and offspring were analyzed with respect to the Trpv6 genotype.
Videomicroscopic Analysis of Sperm Motility-Capacitated spermatozoa were placed in a 100-m chamber at 37°C temperature on a slide warmer and videotaped using a DCR-SR90 Handycam (Sony) connected to an inverted microscope (Axio-Vision 40 CFL, Zeiss). Selected sequences were processed using Adobe PREMIERE software.
Miscellaneous Methods-Whole organ preparation, histological analysis, and Ca 2ϩ measurements in the epididymal fluid were done as described (1).
Statistical Analysis-Data are presented as mean Ϯ S.E. of n independent experiments unless otherwise stated. The Origin 7.0 software (OriginLab) was used for statistical analysis. Significance was assessed with the two-sample Student's t test (p Ͻ 0.05 for significance) unless stated otherwise. Offspring frequency in the mating analysis and weight gain were analyzed using analysis of variance. The occurrence of different genotypes in the inheritance analysis of the Trpv6 Ϫ allele was analyzed using a two-tailed 2 test (GraphPad Software).

Generation and Characterization of Trpv6
Ϫ/Ϫ Mice-We used a Cre-loxP-mediated gene targeting strategy in embryonic stem cells to generate Trpv6-deficient mice (Fig. 1, a and b). We confirmed homologous recombination in Trpv6 ϩ/L2F2 ES cells and Cre-mediated excision of exons 13, 14, and 15, which encode part of transmembrane domain 5, transmembrane domain 6, the pore-forming region in between, and the adjacent amino acids forming the complete intracellular C terminus of TRPV6, in Trpv6 ϩ/Ϫ and Trpv6 Ϫ/Ϫ mice by Southern blot analysis (Fig. 1, c and d). In parallel, mice heterozygous for the Trpv6 L2F2 allele (Trpv6 ϩ/L2F2 ) were bred with FlpeR (129S4/ SvJaeSor-Gt (ROSA)26Sortm1(FLP1) Dym/J) mice (18) to remove the neo r cassette and to produce mice with a conditional Trpv6 ϩ/L2F1 allele (Fig. 1, e and f). Northern blot analysis revealed specific 2.9-kb Trpv6 transcripts expressed in placenta FIGURE 1. Targeted deletion of TRPV6 channel pore and C terminus. We used a Cre-loxP strategy to excise exons 13, 14, and 15, which encode the pore region, part of the fifth and the entire sixth transmembrane-spanning domains, and the cytosolic C terminus of TRPV6. a, the wild-type Trpv6 ϩ allele, targeting construct, and recombinant Trpv6 L2F2 allele. Translated exons (not in scale) are shown as filled boxes. In the Trpv6 L2F2 allele, exons 13, 14, and 15 are flanked by loxP sites (filled triangles). An FRT site (gray triangles)-flanked PGK-neo r cassette is located upstream of the second loxP site. B, BamHI; X, XhoI. Probes and sizes of genomic DNA fragments as expected by Southern blots are indicated. HSVtk, herpes simplex virus thymidin kinase. b, Cre-mediated conversion of the Trpv6 L2F2 allele to the Trpv6 Ϫ/Ϫ allele in mice. c, identification of the recombinant Trpv6 L2F2 allele in recombinant ES cells by Southern blot analysis using 5Ј and 3Ј probes placed externally to the targeted sequence and a neo probe that is directed against the internal PGK-neo r cassette. d, Cre-mediated generation of the Trpv6 Ϫ/Ϫ allele in mice resulted in the conversion of the 3.5-kb fragment of the Trpv6 L2F2 allele to an 11.7-kb fragment (3Ј probe; BamHI digest). The 5Ј probe and XhoI digestion produce a 5.6-kb fragment for the Trpv6 Ϫ allele. In addition, Cre activity also leads to the excision of the PGK-neo r cassette so that the 3.7-kb neo probe signal detectable for the Trpv6 L2F2 allele is absent in Trpv6 ϩ/Ϫ and Trpv6 Ϫ/Ϫ animals (neo probe; XhoI digest). e and f, Flp-mediated generation of the Trpv6 L2F1 allele in mice produces a 5.6-kb fragment (5Ј probe; XhoI digestion) and a 3.5-kb fragment (3Ј probe; BamHI digest) for the Trpv6 L2F1 allele. In addition, Flp activity also leads to the excision of the PGK-neo r cassette so that the 3.7-kb neo probe signal detectable for the Trpv6 L2F2 allele is absent in Trpv6 L2F1 animals (neo probe; XhoI digest). g, Northern blot analysis of poly(A) ϩ RNA isolated from placenta from wild-type (ϩ/ϩ) mice and from epididymis and testis from wild-type (ϩ/ϩ) and Trpv6 Ϫ/Ϫ (Ϫ/Ϫ) mice hybridized with mTrpv6-specific probe). 2.9-kb transcripts were identified in wild-type placenta and epididymis. No signal was detected in wild-type testis or in epididymis and testis of Trpv6 Ϫ/Ϫ mice. Gapdh was used as a loading control. and epididymis but not in testis from wild-type mice (Fig. 1g), supporting our previous finding that TRPV6 is not expressed in testes. After Cre-mediated excision, Trpv6 transcripts were no longer detectable in epididymis of the resulting Trpv6 Ϫ/Ϫ mice (Fig. 1g), confirming effective excision of the Trpv6 gene. However, this lack of TRPV6 expression is in contrast to the results obtained from epididymis of Trpv6 D541A/D541A mice where the identical probe hybridized to Trpv6 transcripts indistinguishable from the transcripts observed in wild type (1). Trpv6 Ϫ/Ϫ mice were viable and showed no obvious anatomical abnormalities, and the Trpv6 Ϫ allele was segregated with the expected Mendelian frequency (Fig. 2a). Analysis of the weight gain of Trpv6 ϩ/Ϫ males and females revealed no growth defects during development and adulthood up to 21 weeks after birth (Fig. 2, b  and c).
Trpv6 Gene Deletion Drastically Diminishes Male Fertility-Breeding analysis revealed that no offspring were born from intercrosses of Trpv6 Ϫ/Ϫ mice in five independent matings (Fig.  2d). To assess whether the impaired reproduction was due to defects in male fertility similar to that in Trpv6 D541A/D541A mice, we performed a systematic breeding analysis with mice of different genotypes. The results showed that Trpv6 Ϫ/Ϫ males did not produce offspring with either Trpv6 Ϫ/Ϫ females or Trpv6 ϩ/ϩ females. In matings of male Trpv6 Ϫ/Ϫ mice with female Trpv6 ϩ/Ϫ mice, only three litters with a total number of only six pups resulted from nine individual matings, whereas matings of male Trpv6 ϩ/Ϫ mice produced 26 litters with altogether 190 offspring in seven matings. The latter frequency corresponds to the frequency of offspring obtained after mating of wild-type mice (1). Apparently, the competence of Trpv6 Ϫ/Ϫ males to fertilize females is considerably reduced. To rule out the possibility that Trpv6 Ϫ/Ϫ males were not copulating, we performed timed matings of wild-type females with either wildtype or Trpv6 Ϫ/Ϫ males (Fig. 2e). We counted the number of plug-positive females and determined that Trpv6 Ϫ/Ϫ males exhibited normal copulatory behavior, but the number of suc-cessfully fertilized oocytes isolated from females impregnated with Trpv6 Ϫ/Ϫ males was negligible compared with wild-type males (Fig. 2f). Accordingly, Trpv6 Ϫ/Ϫ males recognized females and showed normal copulatory behavior but produced pregnancies and offspring with only minor success, indicating that Trpv6 deletion leads to male hypofertility. Impaired Motility and Fertilization Capacity of Sperm from Trpv6 Ϫ/Ϫ Mice-Videomicroscopic analysis of capacitated sperm from wild-type mice (see supplemental Movie S1, wildtype sperm) indicated forceful beating and progressive movement. In contrast, most of the Trpv6 Ϫ/Ϫ sperm were immotile with impaired movement and bending in the tail region (see supplemental Movie S2, Trpv6 Ϫ/Ϫ sperm). Computer-assisted sperm analysis of capacitated sperm isolated from the cauda epididymis revealed that the number of motile sperm from Trpv6 Ϫ/Ϫ mice (Fig. 3a) and their progressive motility were significantly decreased compared with sperm from wild-type mice (Fig. 3, a and b). Even within the small population of Trpv6-deficient sperm that showed progressive motility (Fig. 3, c-e), all speed parameters were significantly reduced, demonstrating that the Trpv6 deletion critically affected sperm motility. The analysis of a second, independent Trpv6 Ϫ/Ϫ mouse line, which we generated from the Trpv6 ϩ/D541A mice (1), confirmed these results (Fig. 3, a-e).
We performed in vitro fertilization experiments to analyze the ability of Trpv6 Ϫ/Ϫ sperm to fertilize oocytes. Therefore, isolated mature eggs from wild-type females were incubated with capacitated wild-type or Trpv6 Ϫ/Ϫ sperms, and the successful fertilization (indicated by the development of two-cellstage embryos) was examined after 24 h. 161 of 238 eggs were fertilized by wild-type sperm, but only 18 of 265 eggs were fertilized by sperm derived from Trpv6 Ϫ/Ϫ mice (Fig. 3f).
The reproductive tracts of wild-type and Trpv6 Ϫ/Ϫ males were macroscopically examined and showed no obvious differences (Fig. 4a). No morphological differences were observed in histological sections of the testes from Trpv6 Ϫ/Ϫ and wild-type

D541A Pore Mutation Inactivates TRPV6 Like Its Deletion
MAY 25, 2012 • VOLUME 287 • NUMBER 22 JOURNAL OF BIOLOGICAL CHEMISTRY 17935 males (Fig. 4b); all cell types that can be observed during spermatogenesis, such as spermatogonia, spermatocytes, and spermatids, were apparent. In sections through the epididymis, we observed that the lumen of caput (Fig. 4c) and cauda (Fig. 4d) epididymis were filled with sperm. Additionally, microscopic analysis with higher magnification demonstrated that there were no differences between the morphology of the epididymal epithelium in Trpv6 Ϫ/Ϫ and wild-type mice (Fig. 4, c and d,  lower panels). Eosin-nigrosin staining of sperm (Fig. 4e) revealed that the number of viable Trpv6 Ϫ/Ϫ sperm isolated from the caput epididymis was 62.53 Ϯ 4.71% (1219 sperm analyzed from four mice) and from the cauda epididymis was 6.66 Ϯ 2.84% (n ϭ 4; 3003 sperm analyzed). In our previous study, we have shown that the reduction of the number of eosin-negative sperm during the epididymal passage was also more than 10-fold in Trpv6 D541A/D541A mice but only 2-fold in wild-type mice (1).
As shown in Fig. 1g, Trpv6 transcripts were not detectable in testes from wild-type mice. Also, no defects in spermatogenesis were observed in the testes of Trpv6 Ϫ/Ϫ mice. Taken together, these results suggest (like in the Trpv6 D541A/D541A mice) a secondary cause of the impaired viability of cauda epididymal sperm in Trpv6 Ϫ/Ϫ mice. We have shown that the TRPV6 and TRPV6 D541A proteins are apparently expressed in the apical membrane of epididymal epithelial cells (1) and the lack of TRPV6 expression in epithelial cells from Trpv6 Ϫ/Ϫ mice (see Fig. 1g and Ref. 1). A luminal Ca 2ϩ gradient along the epididymal duct has been described (14) with luminal Ca 2ϩ being lower in the cauda than in the caput, and we had found that luminal Ca 2ϩ is markedly increased in the cauda from Trpv6 D541A/D541A mice when compared with the Ca 2ϩ levels in wild-type mice. Therefore, we analyzed whether the Ca 2ϩ concentration of the intraluminal fluid in the cauda epididymis in Trpv6 Ϫ/Ϫ mice is affected to the same extent using ion-selective microelectrodes.  Fig. 5a). Next, we measured 45 Ca 2ϩ uptake in the cauda epididymis. Fig. 5b shows a 9-fold decrease in the ratio of 45 Ca/ 51 Cr in the epididymis from Trpv6 Ϫ/Ϫ mice compared with wild-type littermates, indicating a severe defect in Ca 2ϩ absorption.
A higher luminal Ca 2ϩ concentration may result in the formation of Ca 2ϩ precipitates. Hence, alizarin red stainings of epididymal tissue sections were performed, but no Ca 2ϩ depositions were recognized in epididymal sections (Fig. 6a). As we have shown previously (1,5), TRPV6 is also expressed in the prostate (Fig. 6c). In contrast to the findings in the epididymis, the ventral, dorsal, and lateral lobes of the Trpv6 Ϫ/Ϫ prostate (Fig. 6b, lower panel) showed Ca 2ϩ depositions, and Ca 2ϩ precipitates were macroscopically detectable as white rigidification (Fig. 6b, upper panel). The ducts were enlarged and showed a loss of epithelial infoldings into the lumen (Fig. 6b, middle  panel).

DISCUSSION
In this study, we showed that deletion of part of the transmembrane domains including the pore-forming region and the entire C terminus of TRPV6 proteins led to hypofertility in Trpv6 Ϫ/Ϫ males. A systematic breeding analysis revealed that the number of offspring in matings with homozygous Trpv6 Ϫ/Ϫ males was markedly diminished regardless of the genotype of the females, whereas their copulatory behavior as well as the morphology of testis and epididymis was unaffected. However, the motility, fertilization capacity, and viability of sperm derived from cauda epididymis of Trpv6 Ϫ/Ϫ males were drastically reduced; Ca 2ϩ concentrations were abnormally high in the caudal fluid of Trpv6 Ϫ/Ϫ males; and Ca 2ϩ uptake by the epididymal epithelium was drastically impaired. Ca 2ϩ precipitations were not identified in the epididymis but were found in enlarged ducts of the ventral, dorsal, and lateral lobes of the prostate where Trpv6 is also expressed in epithelial cells as in epididymal epithelial cells (1). As in Trpv6 D541A/D541A mice, the

JOURNAL OF BIOLOGICAL CHEMISTRY 17937
Ca 2ϩ deposits in the prostate cannot account for the massive elementary changes observed in caudal sperm of Trpv6 Ϫ/Ϫ mice because these sperm are located upstream of the prostate and were never exposed to prostate secretions at this stage. Nevertheless, it cannot be fully excluded that cell-autonomous alterations in epithelial cells of the prostate contribute at least to some extent to the drastically reduced rate of offspring found in our mating analysis.
All functional deficits in the male reproductive tract that were found in Trpv6 Ϫ/Ϫ mice are virtually identical to those we have found in Trpv6 D541A/D541A mice and to those that have been described (1), suggesting that both the Trpv6 gene excision (this study) and the specific D541A pore mutation (1) lead to a complete inactivation of TRPV6 channels. This conclusion is supported by a detailed comparison of the findings obtained with either mouse line. Development of body weight was not different between Trpv6 Ϫ/Ϫ mice and Trpv6 D541A/D541A mice (Tables 1 and 2). The fertility rate of homozygous mutant males (Trpv6 Ϫ/Ϫ or Trpv6 D541A/D541A ) was assessed in matings with either homozygous (Trpv6 Ϫ/Ϫ and Trpv6 D541A/D541A ) or heterozygous (Trpv6 ϩ/Ϫ and Trpv6 ϩ/D541A ) females and normalized to that obtained using Trpv6 ϩ/D541A or Trpv6 ϩ/Ϫ males, respectively. Apparently, the reduction in fertility rate

TABLE 4 Analysis of copulatory behavior and in vivo fertilization rate
For analysis of copulatory behavior, the ratio of plug-positive wild-type females per mating with male Trpv6 D541A/D541A and Trpv6 Ϫ/Ϫ mice, respectively, was normalized to the ratio obtained from matings of wild-type females with wild-type males.
Corresponding normalization was performed to compare the efficiency of in vivo fertilization assessed by microscopic analysis of flushed embryos from plug-positive wild-type females with males of the indicated genotype. No significant differences were observed between matings with Trpv6 D541A/D541A and Trpv6 Ϫ/Ϫ mice (p Ͼ 0.05). The number of matings for each experiment is indicated (n). was not different between mice with either Trpv6 mutation (Table 3). Also, no differences were observed in copulatory behavior or the rate of embryos isolated from plug-positive females of matings with males homozygous for either Trpv6 mutation (Table 4). Motility parameters of caudal sperm were also reduced to the same extent, similar to the vitality of caudal sperm (Table 5); the only parameter in which we found a difference between sperm from Trpv6 Ϫ/Ϫ and Trpv6 D541A/D541A males was the overall motility, which could be due to an unusually high variability of this parameter in the group of Trpv6 Ϫ/Ϫ mice. Concomitantly, the increase in Ca 2ϩ concentration in the epididymal fluid and the reduction in Ca 2ϩ uptake by the epididymal epithelium were identical in Trpv6 Ϫ/Ϫ and Trpv6 D541A/D541A males (Table 6). Such congruent functional deficits in Trpv6 Ϫ/Ϫ mice and Trpv6 D541A/D541A mice were not necessarily expected because Trpv6 D541A/D541A mice, in contrast to Trpv6 Ϫ/Ϫ mice, express TRPV6 proteins with an intact C terminus (Fig. 7) that contains binding sites for various proteins (19) including S100A10-annexin 2 (22), calmodulin (23), the PDZ protein Na ϩ /H ϩ exchanger regulatory factor 4 (24), and Rab11a (25). These channel-associated proteins are hypothesized to integrate TRPV6 in the cytoskeletal network and to regulate the trafficking, anchoring, and activity of TRPV6 proteins at the plasma membrane, but it is unknown whether the thus formed protein complex has any additional functions. Whereas all these processes may work in Trpv6 D541A/D541A mice, they cannot work in Trpv6 Ϫ/Ϫ mice. Also, in other cation channels, such differences between mutations of the channel pore, even if they have profound effects on conductivity and permeability, and a complete loss of the corresponding channel protein are not unheard of. For instance, the G156S mutation in the pore-forming region of the Kir3.2 potassium channel causes a marked reduction of corresponding K ϩ channels activated by G-protein-coupled receptors sim- FIGURE 7. Model of architecture of wild-type TRPV6 proteins, TRPV6 D541A proteins, and potential proteins produced from the Trpv6-null (Trpv6 ؊ ) allele. a, TRPV6 exhibits the typical topology of all members of the transient receptor potential family with six transmembrane regions and a short domain between transmembrane segments 5 and 6 forming the channel pore. The N-terminal region of wild-type TRPV6 contains at least five (possibly six) ankyrin repeats (5) and a binding site for calmodulin (CaM) (32). In the C-terminal tail of TRPV6, there are binding sites for calmodulin (23), Na ϩ /H ϩ exchanger regulatory factor 4 (NHERF4) (24), Ras-related protein Rab-11A (Rab11a) (25), and S100A10-annexin 2 (22). b, replacement of aspartic acid at position 541 in the pore region of TRPV6 (Asp-541) eliminates Ca 2ϩ conductivity of TRPV6 channels. This TRPV6 D541A pore mutant protein is properly expressed and trafficked to the plasma membrane (1) and contains all binding sites for the TRPV6 interaction partners that anchor TRPV6 channel proteins in its native environment. c, proteins produced in cells homozygous for the Trpv6-null allele (Trpv6 Ϫ ) lack part of transmembrane domain 5, the pore region, transmembrane domain 6, and the entire C-terminal tail due to the genetic ablation of exons 13, 14, and 15. Even if such truncated TRPV6 proteins are stable, the C-terminal binding sites for TRPV6-assocciated proteins are missing.   (26,27), whereas the spontaneous seizure activity observed in Kir3.2 Ϫ/Ϫ mice is not evoked by the Kir3.2 G156S mutation. In another report, the Q618R mutation in the GluR␦2 channel pore that abrogates Ca 2ϩ permeability rescues the deficits in synaptic plasticity of hippocampal neurons and motor coordination that were observed in GluR␦2 Ϫ/Ϫ mice (20,21). We found that the TRPV6 pore mutant (with an intact but non-functional protein still in place) led to very much the same phenotype as the deletion of almost one-third of the gene, which may lead to no TRPV6 protein expression at all or, if truncated proteins are formed, to a TRPV6 protein lacking 222 of 727 amino acid residues. Therefore, we conclude that the main function of TRPV6 proteins is their channel function and that even the persistence of the complete protein TRPV6 D541A does not contribute to functions that are absent in the Trpv6 Ϫ/Ϫ mouse line.

Analysis
In addition to the fertility analyses in the Trpv6 D541A/D541A and Trpv6 Ϫ/Ϫ mouse lines, there are also no differences in their gross appearance including growth. However, they differ significantly from a Trpv6-deficient mouse line generated by Bianco et al. (28). In this mouse line, the genomic sequences including exons 9 -15 of the Trpv6 gene as well as exons 15-18 of the adjacent Ephb6 gene were deleted and replaced by a neomycin resistance cassette. In our Trpv6 Ϫ/Ϫ mouse line, exons 17 and 18 of the Ephb6 gene were deleted, whereas in the Trpv6 D541A/D541A mouse line (1), all exons of the Ephb6 gene were unchanged. Additionally, inactivation of Ephb6 does not affect fertility as shown in two independent Ephb6 Ϫ/Ϫ mouse lines (29,30). The Trpv6 Ϫ/Ϫ mice described by Bianco et al. (28) exhibit significant growth retardation, alopecia in 80% of all mice homozygous for the targeted allele, and impaired fertility in males but also in females (which was not characterized). Differences in growth, hair coat, or fertility of females were not observed in our Trpv6 Ϫ/Ϫ or Trpv6 D541A/D541A mice (this study and Ref. 1). The reason for the observed discrepancy with our Trpv6-deficient mouse line is not known, but differences may be due to deletion of either exons 15-18 of the Ephb6 gene in Trpv6 Ϫ/Ϫ mice described by Bianco et al. (28) or exons 17 and 18 in our mouse line. Additionally, disposition of the promotordriven neomycin resistance cassette, which might affect expression of the adjacent Trpv5 gene, or differences in the genetic background between the mouse lines cannot be excluded. The highly Ca 2ϩ -selective TRPV5 channels resemble TRPV6 in many aspects as they are also constitutively active and exhibit many features possessed by Ca 2ϩ transporters in epithelial cells (6,7). TRPV5 plays a critical role in Ca 2ϩ reabsorption in the kidney epithelium of collecting ducts, but Trpv5 Ϫ/Ϫ mice show no defect in fertility (31). Obviously, the defect in Ca 2ϩ uptake in the epididymal epithelium induced by deletion of the Trpv6 gene could not be compensated for by the nearest relative, TRPV5, demonstrating that TRPV6 is essential for the posttesticular sperm maturation process.
In summary, we demonstrated that Trpv6 Ϫ/Ϫ mice show phenotypic changes in the male reproductive tract equal to those in Trpv6 D541A/D541A mice, corroborating the crucial role of TRPV6 in the epididymis for the development of fertilization capacity of sperm. We conclude that TRPV6 proteins are essential components of Ca 2ϩ -conducting channel complexes in the apical membrane of epididymal epithelial cells and are responsible for decreasing the Ca 2ϩ concentration of the intraluminal fluid in the cauda epididymis. Furthermore, the results obtained with the Trpv6 Ϫ/Ϫ mice underscore the finding in Trpv6 D541A/D541A mice that appropriate regulation of intraluminal Ca 2ϩ concentration in the epididymal duct is essential for the production of fertilization-ready spermatozoa during the epididymal passage. Together with the phenotype analysis of Trpv6 D541A/D541A mice, our results argue for the conclusion that the D541A pore mutation leads to a complete inactivation of TRPV6 channels in the epididymal epithelium.