Plasma Membrane Ca2+ ATPase 4 Is Required for Sperm Motility and Male Fertility*

Calcium and Ca2+-dependent signals play a crucial role in sperm motility and mammalian fertilization, but the molecules and mechanisms underlying these Ca2+-dependent pathways are incompletely understood. Here we show that homozygous male mice with a targeted gene deletion of isoform 4 of the plasma membrane calcium/calmodulin-dependent calcium ATPase (PMCA), which is highly enriched in the sperm tail, are infertile due to severely impaired sperm motility. Furthermore, the PMCA inhibitor 5-(and-6)-carboxyeosin diacetate succinimidyl ester reduced sperm motility in wild-type animals, thus mimicking the effects of PMCA4 deficiency on sperm motility and supporting the hypothesis of a pivotal role of the PMCA4 on the regulation of sperm function and intracellular Ca2+ levels.

Calcium and Ca 2؉ -dependent signals play a crucial role in sperm motility and mammalian fertilization, but the molecules and mechanisms underlying these Ca 2؉dependent pathways are incompletely understood. Here we show that homozygous male mice with a targeted gene deletion of isoform 4 of the plasma membrane calcium/ calmodulin-dependent calcium ATPase (PMCA), which is highly enriched in the sperm tail, are infertile due to severely impaired sperm motility. Furthermore, the PMCA inhibitor 5-(and-6)-carboxyeosin diacetate succinimidyl ester reduced sperm motility in wild-type animals, thus mimicking the effects of PMCA4 deficiency on sperm motility and supporting the hypothesis of a pivotal role of the PMCA4 on the regulation of sperm function and intracellular Ca 2؉ levels.
Successful fertilization requires the sperm to travel long distances and undergo capacitation prior to reaching the female egg. After reaching their target, the sperm must interact with the extracellular matrix of the egg, including proteins of the zona pellucida, and release acrosomal material. Calcium is considered to exert a function on most, if not all, of these processes. In this field, most of the work on Ca 2ϩ signaling has focused on Ca 2ϩ entry mechanisms, especially on the role of Ca 2ϩ channels (1)(2)(3)(4). For example, gene ablation of the cation channel of sperm (CatSper) leads to impaired sperm motility and male infertility (5), and mice lacking the mitochondrial voltage-dependent anion channel type 3 (VDAC3) are also infertile due to immotile sperm (6). These results show that tight regulation of ion entry by ion channels is critical to sperm function. Although there is little doubt as to the importance of calcium homeostasis in sperm motility and fertilization (7)(8)(9)(10)(11)(12), the function of the plasma membrane Ca 2ϩ /calmodulin-dependent Ca 2ϩ ATPase (PMCA) 1 during this process remained enigmatic.
PMCA represents a family of enzymes that extrude calcium from the cytosol across the plasma membrane of eukaryotic cells. Since their initial identification in erythrocytes (13), four different isoforms have been identified, and multiple splice forms of these isoforms have been described. The well defined tissue-specific expression pattern of different isoforms and splice variants of the pump in various mammalian tissues (14) and the regulated expression pattern during mouse development (15) strongly suggest a specific physiological function for each isoform and splice variant (reviewed in Strehler and Zacharias (16)). The identification of physical and functional interaction partners of the Ca 2ϩ pump has given insights into the putative functions of PMCAs as regulators of Ca 2ϩ -dependent signal transduction processes (17)(18)(19)(20)(21). Interaction of PMCA2 and -4 "b" splice variants was shown to be mediated by the PDZ-(PSD-95/Dlg/ZO-1) domain of the corresponding interaction partner and the C termini of the PMCA isoform (which harbors a typical PDZ domain binding motif (17)). Both modes of interaction with PDZ domain-containing proteins, specific and promiscuous binding to different PDZ domains, have been demonstrated (18,19). In addition to the overlapping expression pattern of the four PMCA isoforms and the diversity generated by alternative splicing, the specificity of interaction with other proteins adds a further level of complexity in determining the physiological functions of each isoform.
Gene ablation in mice using homologous recombination in embryonic stem cells represents one possibility to evaluate the function of proteins in vivo and to address the isoform-specific functions of a certain protein. This strategy has been successfully used to generate PMCA2-deficient mice that suffer from deafness and balance deficits (22), supported by analyses of "deafwaddler" and "wriggle mouse Sagami" mouse strains, both showing a phenotype comparable with the PMCA2-deficient mice and also harboring spontaneous mutations in the PMCA2 gene (23,24).
To clarify the in vivo function of PMCA4, we generated PMCA4-deficient mice and studied the physiological effects of this gene deficiency. PMCA4 deficiency does not impair development to adulthood but leads to male infertility due to impaired sperm motility. * This work was supported by Grant SFB 355 from the Deutsche Forschungsgemeinschaft, an MRC International Appointee grant (to L. N.), and a grant from the Novartis Foundation, Germany. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AY560895 and AY560896.

EXPERIMENTAL PROCEDURES
Cloning of Mouse PMCA4 Isoforms-Mouse PMCA4 isoforms were cloned by rapid amplification of cDNA ends from a mouse testis Marathon-Ready TM double-stranded cDNA library (catalog number 7455-1, Clontech) using the Advantage® 2 PCR system (Clontech) and the following gene-specific primers: mouse PMCA4 forward, 5Ј-GTC TGA TCA TGT CTG TCC TCA CAG TTG-3Ј, and mouse PMCA4 reverse, 5Ј-GCA GCC CCT CTG GCA CAG CCA CT-3Ј. Polymerase chain reaction was performed as suggested by the manufacturer, and the resulting PCR fragments were cloned into pCR®-XL-TOPO vector (Invitrogen) and subsequently sequenced with the standard T7 and M13 reverse primers. The resulting sequences were analyzed and aligned with DNAMAN 4.0 software (Lynnon BioSoft), and the transmembrane helices were predicted with HMMTOP Version 2.0 at www.enzim.hu/ hummtop (25). Mouse PMCA4b and -4a sequences have been deposited in GenBank TM , and the accession numbers are AY560895 and AY560896, respectively.
Northern Blotting and Immunofluorescence Stainings-To determine the expression pattern of PMCA4 in mouse testis, we hybridized the MessageMap TM Northern blot (each lane containing 2 g of poly(A)ϩ RNA, Stratagene, catalog number 776900) with the 32 P-dATPlabeled full-length PMCA4b cDNA according to the manufacturer's protocol and reprobed the subsequently stripped membrane with a 32 P-dATP-labeled ␤-actin probe. The subcellular distribution of PMCA4b was determined by immunofluorescence stainings. Isolated sperm were allowed to swim out from cauda epididymidis in sperm preparation buffer (MediCult, catalog number 10680060) for 15 min, and the suspension was streaked out on poly-lysine-coated slides, airdried for 10 min, and fixed and stained as described previously (21). The PMCA4-specific polyclonal antiserum was also described previously (26).
Generation of PMCA4-deficient Mice-A 12-kb genomic DNA fragment containing exons 2-4 of the mouse PMCA4 gene was isolated from a 129Sv mouse genomic library in the Fix TM II vector (Stratagene). The targeting vector was generated by inserting the homologous BglII/KpnI and BspMII/NotI fragments into the pPNT vector (27) (Fig. 3A). 20 g of linearized targeting vector was electroporated into E14.1 embryonic stem cells. Neomycin-and ganciclovir-resistant clones (500 g/ml G418 and 2 M ganciclovir) were screened for homologous recombination by Southern blot using the external BamHI/BglII fragment of the mouse PMCA4 gene as a probe (Fig. 3A). This probe hybridizes to a 7.0-kb BamHI fragment (endogenous pmca4 allele) and a 3.3-kb fragment (disrupted allele). Embryonic stem cells that had undergone homologous recombination were injected into C57Bl/6 blastocysts, and the manipulated blastocysts were transferred into pseudo-pregnant foster mice to generate chimeras. Chimeric males were mated to C57Bl/6 females to test for germ line transmission of the targeted PMCA4 allele. PMCA4-deficient mice were obtained by appropriate inbreeding of heterozygous offspring and subsequent genotyping.
RT-PCR and Western Blotting-The absence of PMCA4 mRNA in PMCA4-deficient mice was tested by RT-PCR using the OneStep RT-PCR kit (Qiagen) and the following PMCA4-specific (not species-specific) primers: PMCA4 forward, 5Ј-CTG AGG AAG CTC ATGGAG C-3Ј, and PMCA4 reverse, 5Ј-CGG AAA/G TGC TTC TCT TTG C-3Ј. To test for the absence of PMCA4 at the protein level in sperm, isolated sperm from one cauda epididymidis were collected by centrifugation and boiled for 5 min in 200 l of Laemmli sample buffer (Bio-Rad). 20 l of each lysate was separated on a 10% SDS-PAGE, blotted onto a nitrocellulose membrane (Schleicher & Schuell), and the membranes were cut at the level of the 75-kDa protein marker band. The upper half was tested for PMCA4 expression using the PMCA4-specific monoclonal antibody JA9 (28) (NeoMarkers, 1:500 dilution), and the lower half was probed for actin (polyclonal goat anti-actin, Santa Cruz Biotechnology, catalog number sc-1616, 1:500 dilution) to check for equal loading of the samples. Blocking and antibody incubations were done with 5% milk in phosphate-buffered saline/0.05% Tween 20. Signals were detected with specific horseradish peroxidase-labeled secondary antibodies and the ECL detection system (Amersham Biosciences).
Tissue Sections of Testis and Histological Stainings-Tissue sections were frozen and prepared as described previously (21) and stained with hematoxylin and eosin using a standard protocol. Cytological staining of mouse sperm was carried out using SpermacStain TM (Stain Enterprises Inc.) according to the manufacturer's instructions.
In Vitro Fertilization-Sperm were collected from cauda epididymidis in sperm preparation buffer (MediCult) and capacitated in vitro for 2 h at 37°C. Oocytes were prepared from C57Bl/6 females that had been synchronized with 10 units of pregnant mare serum gonadotropin (Sigma) and 10 units of human chorionic gonadotropin (Sigma) 48 and 14 h prior to oocyte collection. Eggs were flushed from oviducts in M2 medium (Sigma) and cultivated in M16 medium (Sigma) in 5% CO 2 at 37°C. In vitro fertilization capacity was tested as described previously (29). In brief, eggs were incubated with ϳ10 5 wild-type or 10 5 PMCA4deficient sperm for 24 h at 37°C, and eggs that had divided to the two-cell stage were counted as indicative of successful fertilization.
Estimation of Sperm Motility and Measurement of Intracellular Calcium-Sperm were collected from cauda epididymidis as described above. The supernatant containing sperm was decanted into a fresh tube, and the cells were left untreated at 37°C for 30 min or loaded with the PMCA inhibitor 5-(and 6)-carboxyeosin diacetate succinimidyl ester (10 M, Molecular Probes).
Overall sperm motility was estimated using the medeaLAB CASA 4.2 system (Erlangen) optimized for mouse sperm. Morphology and tracking threshold upper levels were set to: red 255, green 90, and blue 255, respectively. Additionally, all possible form parameters were deactivated (e.g. flagellum detection, color, and others), and area and form factor filters have been left unchanged. In addition to the classification of sperm motility, average path velocity, progressive velocity, and track speed were calculated with statistical analysis of raw motility data.
To estimate intracellular Ca 2ϩ levels in capacitated sperm, they were prepared in sperm preparation buffer (MediCult) as described above, capacitated for 30 min, and loaded with 10 M Fluo-4-AM for 30 min at 37°C. The cells were subsequently washed, counted (Coulter counter), and resuspended at a concentration of 1 ϫ 10 7 cells/ml in sperm prep-

RESULTS AND DISCUSSION
Full-length cDNA of mouse PMCA4b splice variant and the C terminus of mouse PMCA4a variant were cloned from a testis cDNA library by rapid amplification of cDNA ends (5Ј and 3Ј rapid amplification of cDNA ends. Comparison of predicted protein sequences with human (30) and rat (31) PMCA4b and -4a revealed a high degree of homology (Fig. 1). Mouse PMCA4b contains 10 predicted transmembrane domains, forming a pore in the plasma membrane, which is a typical feature of PMCAs. The C terminus of mouse PMCA4b harbors a typical PDZ domain binding motif (amino acid sequence: . . . ETPV), most presumably mediating specificity of binding to certain PDZ domains, as shown previously for other human PMCA b splice variants (17)(18)(19)(20)(21)32).
The C terminus of mouse PMCA4a has a high level of similarity to "a" splice variants of other species (Fig. 1;  However, the molecular function of this C-terminal PMCA4a motif remains unclear. Although previously shown to be expressed in several organs (14), multiple tissue Northern blot analysis revealed prominent expression of the PMCA4 messenger RNA in mouse testis ( Fig.  2A). In addition, although the relative actin mRNA contents of various tissue types are most likely different, the prominent PMCA4 signal in Fig. 2A suggests robust expression of PMCA4 in mouse testis.
A polyclonal antibody directed against the N-terminal part of PMCA4 and cross-reactive with mouse PMCA4 (26) was used to determine subcellular localization of PMCA4 protein in wildtype mouse sperm. The protein was expressed in the principal piece of the sperm tail, the flagellar apparatus propelling the spermatozoon forward, and to a lesser extent to the sickleshaped mouse sperm acrosome region (Fig. 2B). Interestingly, the recently described sperm-specific calcium channel CatSper was also localized in the principal region of the tail, and its gene deletion leads to severely reduced sperm motility and male infertility, thus underlining the important role of Ca 2ϩ signaling in sperm motility (5).
To study the physiological function of PMCA4 in vivo, we disrupted the PMCA4 gene in embryonic stem cells by homologous recombination (Fig. 3A). The second exon and part of the third exon were replaced by the neomycin resistance cassette. Following homologous recombination, embryonic stem cells were injected into blastocysts and implanted into pseudo-preg- nant foster mice. Following germ line transmission of the mutation, PMCA4-deficient mice were obtained by cross-breeding heterozygous offspring. Disruption of the PMCA4 gene was confirmed by Southern blotting (Fig. 3B), and the absence of the mRNA transcript and of the protein was shown by RT-PCR and Western blot analysis (Fig. 3, C and D).
Offspring of mated heterozygous males and females were born in the expected Mendelian ratio (26.2% ϩ/ϩ, 46.3% ϩ/Ϫ, 27.5% Ϫ/Ϫ), suggesting that PMCA4 deficiency did not affect embryonic development. PMCA4 Ϫ/Ϫ mice were indistinguishable from their wild-type littermates with respect to body weight, appearance, and gross behavior. Adult PMCA4 Ϫ/Ϫ females, mated with wild-type or heterozygous PMCA4 ϩ/Ϫ males, did not show alterations in fertility (100% fertile). However, a homozygous PMCA4-deficient line could not be established when both homozygous males and females were crossed. Appropriate homo-/heterozygote cross-breeding demonstrated normal female and absent male fertility; 10 PMCA4 Ϫ/Ϫ males engendered no pregnancies over a period of up to 6 months. Alterations in mating behavior or erectile dysfunction were excluded because after mating homozygous knock-out males with PMCA4-deficient females, the latter had a normal frequency of vaginal sperm plugs.
A closer microscopical examination of testes and sperm revealed no histological differences in testes architecture and no morphological differences in sperm of PMCA4 Ϫ/Ϫ mice and their wild-type littermates (Fig. 4, A and B). In vitro fertilization assays were performed to test the ability of PMCA4-deficient sperm to fertilize eggs. 38% (10 of 24) of eggs incubated with capacitated wild-type sperm and 35% (11 of 30) of eggs incubated with PMCA4-deficient sperm reached the two-cellstage after 24 h (example in Fig. 4C). PMCA4-deficient sperm were also able to bind to empty zona pellucida, suggesting that these sperm undergo the normal acrosome reaction (example in Fig. 4C). To gain a first insight into the regulation of intracellular calcium of PMCA4-deficient sperm after capacitation, we have estimated the intracellular calcium after preparation and 60-min capacitation of sperm from caudae epididymides from PMCA4-deficient mice in comparison with sperm of their wildtype littermates. Assuming a K D (Ca 2ϩ ) of 345 nM for Fluo4, the average intracellular calcium concentration of wild-type sperm was found to be 157 nM, and the intracellular calcium in knockout sperm was 370 nM (example of one recording given in Fig.  4D, in total n ϭ 15, p Ͻ 0.05). This underlines the previously suggested pivotal role of the PMCA in the regulation of basal calcium levels and calcium clearance in sperm (33), but detailed analyses of the regulation of intracellular free calcium in different compartments of sperm have to be made to understand the functions of the PMCA4 in the fertilization process.
As a consequence of the genetic manipulation, an obvious difference in the motility of mutant and wild-type sperm was observed. Sperm were classified by standard clinical tests for sperm motility disorders with a computer-aided sperm analysis system; a large number of PMCA4-deficient sperm were immotile (68%), displayed extremely low directed progressive motility (14%), or showed no directed movement (18%) as compared with sperm of wild-type littermates (7% immotile, 73% progressive motility, 20% no directed motility, examples of recordings in Fig. 5, A and B). Analysis of main motility parameters showed that average path velocity, progressive velocity, and track speed were severely impaired in PMCA4-deficient mice (Fig. 5D).
If deletion of PMCA4 leads to strongly impaired sperm motility, it should conceptually be possible to mimic this effect by FIG. 4. Morphology and function of PMCA4 ؊/؊ testis and PMCA4-deficient sperm. A, hematoxylin/eosin staining of testis sections did not show histological differences between PMCA4 Ϫ/Ϫ and wild-type (WT) littermate controls. KO, knock-out. B, no obvious cytological differences were observed in bright field microscopy of previously stained, isolated sperm (ϫ63 objective, oil immersion, SpermacStain® staining). C, no significant differences in in vitro fertilization capacity of wild-type (upper panel) and PMCA4-deficient sperm (middle panel) were observed. PMCA4-deficient sperm were able to attach to empty zona pellucida, suggesting normal acrosome reaction (lower panel). D, example of one recording of change in intracellular calcium in response to 50 M Ca 2ϩ ionophore A23187. PMCA4-deficient sperm (red) showed an elevated resting intracellular Ca 2ϩ concentration [Ca 2ϩ ] i and approximately the same maximal Ca 2ϩ concentration in response to ionophore A23187 as the wild-type control sperm (blue). To estimate intracellular calcium, similar recordings in the presence of EDTA have been made (not shown), and the intracellular free calcium levels have been calculated. Time in seconds, relative intensity ϭ relative intensity of fluorescence.
using the cell-permeable PMCA inhibitor 5-(and-6)-carboxyeosin diacetate succinimidyl ester (CE). 10 M CE severely reduced the motility of wild-type sperm (36% immotile, 30% progressive motility, 34% no directed motility, examples of recordings in Fig. 5C, analysis of main motility parameters in Fig. 5D), providing additional evidence for the importance of PMCA activity in the regulation of sperm motility. CE in the concentration range of 10 -20 M has been shown previously to be an effective, specific inhibitor of the plasma membrane calcium pump (34 -36).
In conclusion, we describe a highly specific form of male infertility in animals with a deletion of the PMCA4 gene. The latter is crucial to sperm motility but not to in vitro fertilization capacity. The PMCA inhibitor 5-(and-6)-carboxyeosin diacetate succinimidyl ester mimics the effect of gene deletion on sperm motility. Therefore, it may act as an ideal lead compound for the development of contraceptive drugs and has a number of chemically active side groups that can potentially be used for modification, for example to enhance accumulation in the testis or seminal vesicles and reduce potential side effects. Little is known about the toxicity of 5-(and-6)-carboxyeosin diacetate succinimidyl ester, but the parent compounds eosin and fluorescein are extensively used as a colorant in cosmetics and for diagnostic purposes, respectively, and therefore, they represent a promising class of lead compounds for further contraceptive drug development. Conceptually, our findings may open the way to a novel form of non-hormonal contraception. Furthermore, our findings make PMCA4 a candidate gene for the analysis of genetic and environmental causes of male infertility, an increasing problem globally (37,38).