Absence of Dp71 in mdx^3cv mouse spermatozoa alters flagellar morphology and the distribution of ion channels and nNOS

Dystrophin is a member of the protein family encoded by the Duchenne muscular dystrophy (DMD) gene, which is expressed in muscular and non-muscular tissues. The DMD gene has internal promoters (Winder, 1997) that encode short dystrophin products of 260 kDa (Dp260), 140 kDa (Dp140), 116 kDa (Dp116) and 71 kDa (Dp71). Full-length dystrophin and all dystrophin proteins (Dps) are expressed in neural tissue (Blake et al., 2001) and Dp71 is widely distributed in nonmuscle tissues (Lederfein et al., 1993). In addition, by alternative splicing of exons 71-74 and/or 78, several isoforms may be generated (Austin et al., 2000). We have shown the expression of Dp71 in brain subcellular fractions (Chávez et al., 2000) and in spermatozoa (Hernández-González et al., 2001). Dystrophin links cytoskeletal actin to the extracellular matrix via a dystrophin glycoprotein complex composed of dystrophin and dystrophin-associated proteins (DAPs) (Ibraghimov-Beskrovnaya et al., 1992) and builds the DAPC, which is composed of β-dystroglycan, sarcoglycans, dystrobrevins and syntrophins (Ervasti and Campbell, 1993). At the neuromuscular junction and in non-muscular tissues, utrophin, a dystrophin-related protein (DRP), is also associated to DAPs (Clerk et al., 1993). Alternative promoters and alternative splicing give rise to the utrophin protein family: full-size utrophin (400 kDa), DRP-1 (116 kDa) and Up71 (70 kDa), which have different tissue localizations (Wilson et al., 1999). Utrophin has an N-terminal actin-binding domain, which links the actin cytoskeleton to the plasma membrane. From the functional point of view, the C-terminus of both protein families comprises several domains for DAP binding (Winder, 1997).

On the cytoplasmic side of the DAPC, dystrophin binds to dystrobrevins providing a scaffold for binding to syntrophins (α, β₁, β₂, γ₁, γ₂), modular proteins that link ion channels, aquaporin channels and signaling proteins to the C-terminus of dystrophins, utrophins and dystrobrevins (Yang et al., 1995). Dp71 and DAPs are also expressed in non-muscular tissues including the central nervous system (Dalloz et al., 2001), kidney, liver (Loh et al., 2001) and spermatozoa (Hernández-González et al., 2001). Dp71 isoforms and DAPs are localized in specific domains of mammalian spermatozoa and, interestingly, they only express a product of the DMD gene (Hernández-González et al., 2001). Protein components of the DAPC show different localizations in mammalian spermatozoa, α-syntrophin was located in the middle piece of both plasma membrane and flagellum, whereas β-dystroglycan was only located in the plasma membrane of the flagellar middle piece (Hernández-González et al., 2001). Together, Dp71 and the F-actin cytoskeleton, through the α-syntrophin PDZ domain, can anchor different ionic channels and signaling proteins to specific domains of the plasma membrane, named syntrophin-associated proteins (SAPs) (Fig. 1). Some proteins containing the PDZ ligand domain have been found in mammalian spermatozoa such as: K⁺ channels (Félix et al., 2002), Ca²⁺ channels (Darszon et al., 1999), aquaporin-7 and -8 (Calamita et al., 2001) and neural nitric oxide synthase (nNOS) (Hernández-González et al., 2001) (Fig. 1). These molecules are involved in functional processes such as capacitation, acrosome reaction and motility. It was recently reported that the absence of Dp71 in mdx^3cv and Dp71 null mice, disrupts the distribution of the Kir4.1 potassium channels in Müller glial cells without altering their expression (Connors and Kofuji, 2002; Dalloz et al., 2003). Therefore, the aim of the present investigation was to determine whether the absence of Dp71 also alters the distribution of nNOS and ion channels in dystrophic mdx^3cv spermatozoa.

Fertility alterations in DMD patients have not been reported. However, a low reproductive rate has been observed in mdx^3cv mice (Cox et al., 1993). To understand how the absence of Dp71 affects the DAPC in spermatozoa, we analyzed the localization and expression of β-dystroglycan and α-syntrophin, in the dystrophic mutant mdx^3cv and in wild-type mice. Our results indicate that these proteins were affected in mdx^3cv mice. Furthermore, in the absence of Dp71, altered flagellar morphology increased, originating a new DMD phenotype. These abnormalities were partially corrected when the plasma membrane was removed with neutral detergents. Interestingly, Na⁺ and K⁺ channels and the presence and localization of nNOS were disrupted in mdx^3cv spermatozoa. Moreover, a short utrophin product, Up71 was relocalized and upregulated in mdx^3cv spermatozoa. These results indicate that Dp71 and its DAPC are essential components of the sperm functional protein scaffold, necessary for the anchorage of membrane proteins in specific sperm domains and for normal mammalian spermatozoa morphology.

All reagents were of analytical quality and purchased from Sigma (St Louis, MO). Nitrocellulose membranes, acrylamide and N,N′-methylene-bis-acrylamide and molecular weight markers were from Bio-Rad (Richmond, CA).

JAF antibody was produced by D. Mornet and characterized by Rivier et al. (Rivier et al., 1999). Dys2 and DRP-1 antibodies against dystrophins and utrophins respectively, were purchased from Novocastra (Newcastle-upon-Tyne, UK) and nNOS type 1 from BD Transduction Laboratories (Lexington, NY). β-Actin and β-tubulin antibodies were from Sigma (St Louis, MO). Anti K⁺ channel Kv1.1 antibody and its blocking peptide (sc-1118P) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-Na⁺ channel antibody was produced and characterized by Trimmer et al. (Trimmer et al., 1989). Peroxidase HRP-labeled anti-rabbit IgG, HRP-labeled anti-mouse IgG, TRITC (tetramethyl rhodamine)-labeled anti-rabbit IgG and TRITC-labeled anti-mouse IgG were purchased from Jackson ImmunoResearch (West Grove, PA).

C57BL/6 and heterozygote mdx^3cv mutant mice strains were obtained from Jackson Lab. (Bar Harbor, ME), and were bred in our laboratory facilities. They were identified by western blot analysis of brain extracts. All experiments complied with the international regulations for animal care and experimentation.

Cauda epididymal mouse sperm were collected from wild-type mice and mdx^3cv retired male breeders by placing minced cauda epididymis in 1 ml modified Krebs-Ringer medium (Whitten's/HEPES-buffered). After 10 minutes, sperm in suspension were washed in 10 ml Krebs-Ringer medium by centrifugation at 600 g for 5 minutes. Sperm cells were pelleted and then resuspended to a final concentration of 30×10⁶ cells/ml and simultaneously processed for indirect immunofluorescence and western blotting. Aliquots for morphological analysis were fixed in 1.5% final concentration of formaldehyde (Hernández-González et al., 2001).

Sperm suspension samples were fixed in formaldehyde for 60 minutes at room temperature and centrifuged at 600 g for 4 minutes, washed, permeabilized and immunostained with appropriate antibodies, following a previously published protocol (Hernández-González et al., 2001). Spermatozoa were permeabilized in absolute acetone for 7 minutes at –20°C and washed three times with PBS. Specific primary antibodies were added to the spermatozoa samples, incubated for 2 hours at 37°C, and then incubated with the respective TRITC-conjugated secondary antibody. Samples were mounted in PBS-glycerol (1:1) and examined with an epifluorescence microscope (Hernández-González et al., 2001). As negative controls sperm samples were treated with a non-specific first antibody; with only the secondary antibody; or with the anti-Kv1.1 antibody pre-incubated with the blocking peptide. All were then used for immunolocalization studies and western blotting. All controls gave negative results (data not shown).

Wild-type and mdx^3cv spermatozoa suspended in Krebs-Ringer medium were treated with the non-ionic detergent Brij-36T at 1.5% final concentration (Juárez-Mosqueda and Mújica, 1999), in the presence of protease inhibitors (4 mM PMSF, 4 mM pHMB, 2 mM pAB, 2 mM benzamidine, 1 μM leupeptin, 1 μM aprotinin and 1 μM pepstatin) for 5 minutes at 4°C. This treatment solubilized sperm membranes and acrosomes. Immediately, the spermatozoa were centrifuged at 2000 g for 15 minutes at 4°C and the pellet (demembranated spermatozoa) was recovered. The spermatozoa were washed three times with buffer A (50 mM Tris-HCl, pH 7.4, 2 mM EGTA) and collected by centrifugation (2000 g for 15 minutes at 4°C). The final pellet, containing demembranated spermatozoa, was fixed in formaldehyde, as above.

Spermatozoa from wild-type and mdx^3cv mice were resuspended in buffer A with the protease inhibitor mixture and sonicated for 15 seconds at 40 watts (Ultrasonic Processor, Daigger Co., Vernon Hills, IL). This process breaks sperm membranes (plasma and acrosomal) and separates the heads from the flagella. Flagella were then collected by centrifugation at 1000 g for 30 minutes, at 4°C. The supernatant was saved for membrane isolation and the pellet was resuspended in buffer A and layered onto a two-step Percoll gradient (70/95%) and centrifuged for 20 minutes at 600 g at 4°C. The bottom pellet, containing heads without acrosome was discarded and the flagellar fraction was collected from the Percoll interface. Then, the flagellar fractions were resuspended in buffer A plus protease inhibitor mixture, and solubilized by DTT-SDS treatment (Júarez-Mosqueda and Mújica, 1999). Sperm membranes (plasma and acrosomal membranes) were recovered from the 1000 g supernatant and centrifuged at 100,000 g for 60 minutes at 4°C. The pellet was solubilized in SDS (Hernández-González et al., 2000). Proteins from the solubilized sperm fractions were separated by electrophoresis and processed for western blotting.

Spermatozoa (50×10⁶ cell/ml), isolated sperm membranes and isolated flagella, resuspended in 50 mM Tris-HCl and 2 mM EDTA, pH 7.4, containing the protease inhibitor mixture, were treated with 1% SDS (final concentration) and were then boiled for 5 minutes and centrifuged at 10,000 g for 5 minutes. The supernatants were used for total protein determination and processed for electrophoresis and western blotting.

Sperm membranes and flagellar fractions were analyzed by SDS-PAGE (using 10% or 6% acrylamide) and transferred to nitrocellulose membranes. For detection of specific proteins, the nitrocellulose membranes were incubated for 2 hours at 37°C with specific antibodies and then with an appropriate secondary antibody coupled to HRP, developed by an ECL chemiluminescence kit (Amersham) and recorded on film (X-Omat, Kodak) (Hernández-González et al., 2001).

Whole and Brij-treated spermatozoa from wild-type and mdx^3cv mice were fixed in formaldehyde (1.5% final concentration). Sperm morphology was examined by phase-contrast microscopy and classified as follows: (1) with normal flagella, not curved or twisted; (2) with curved flagella in the middle piece; and (3) with twisted flagella. To evaluate flagellar alterations, 1000 spermatozoa (n=5) from each mouse strain were classified and counted. The data were analyzed by Student's t-test using the Sigma-Stat software.

To determine possible morphologic alterations of spermatozoa from dystrophic mice, we compared mdx^3cv and wild-type spermatozoa cell shape. Major changes were only found in the flagellum, three types of flagella were observed: normal (Fig. 2A1), and two altered morphologies: one with curved (or bent) flagella at the middle piece level (Fig. 2A2) and another with twisted flagella (Fig. 2A3). These flagellar morphologies were found in different proportions in sperm samples from wild-type or mdx^3cv mice. Evaluation by optical microscopy (1000 sperm cells, n=5) of wild-type spermatozoa showed that 84% presented normal flagellar morphology and only 16% had anomalous flagellar: 10% were sperm with bent flagella and 6% had twisted flagella (Fig. 2B, black bars). In contrast, 46% of mdx^3cv spermatozoa showed normal morphology, 46% presented curved flagella and 8% exhibited a twisted form (Fig. 2B, white bars). When spermatozoa were treated with Brij-36T, a neutral detergent that solubilizes the sperm membranes and acrosome (Juárez-Mozqueda and Mújica, 1999), most sperm cells from mdx^3cv recovered their normal morphology (91%) and only 8% showed curved flagella (Fig. 2C). The ratio of normal to abnormal flagella from demembranated spermatozoa is similar in both strains (Fig. 2C). These morphological alterations are a new phenotype and the Dp71∼DAPC could be related to these alterations, it was therefore investigated in wild-type and mdx^3cv spermatozoa.

To examine the level and localization of Dp71 in wild-type and mdx^3cv mice spermatozoa, indirect immunofluorescence and western blotting analyses were performed. In spermatozoa from wild-type mice, immunoreactivity of Dys-2 antibody was concentrated as granular staining in the middle piece and in the postacrosomal region (Fig. 3A1). Staining was undetectable in spermatozoa from mdx^3cv mice (Fig. 3A2), confirming the genotype of this dystrophic animal model. Again, we observed the remarkable difference between the morphology of spermatozoa from dystrophic animals and from wild-type mice in the phase contrast images (see Fig. 2A).

We characterized the presence of dystrophins by western blotting of the sperm membrane fraction extracts from wild-type and mdx^3cv mice. Only one protein band with apparent molecular mass 76 kDa was clearly identified in wild-type sperm cell membranes, which corresponded to Dp71 and was undetectable in mdx^3cv spermatozoa membrane fractions (Fig. 3B). No other Dps were detected in spermatozoa protein extracts. As expected, the mdx^3cv mutation abolished the presence of Dp71 as demonstrated here.

In order to determine if some DRP can compensate for the decreased expression or absence of Dp71 in mdx^3cv, we analyzed the concentration and distribution of utrophins in wild-type and mdx^3cv spermatozoa. Fig. 4 compares the localization and presence of utrophins in whole spermatozoa. DRP-1, anti-utrophin antibody, exclusively stained the postacrosomal region of whole spermatozoa from wild-type mice (Fig. 4A1). Analysis of mdx^3cv spermatozoa revealed fluorescence in the middle piece and more intense staining in the whole head (Fig. 4A2). Fluorescence was also more evident in dystrophic compared to control spermatozoa. In addition, utrophin concentration was analyzed by western blotting using purified membranes from wild-type and mdx^3cv mice and as a control for protein concentration, β-actin was detected simultaneously. Fig. 4B shows that the DRP-1 antibody revealed only one protein band of 76 kDa corresponding to Up71. The level of Up71 was higher in membrane fractions of mdx^3cv spermatozoa compared to wild-type spermatozoa, confirming the more concentrated Up71 fluorescence observed by indirect immunofluorescence. No other utrophin proteins were detected in spermatozoa extracts. Therefore, in the absence of Dp71, Up71 was upregulated and redistributed, from the postacrosomal region to the middle piece and the head of spermatozoa.

We previously demonstrated that non-muscle cells express some members of the DAPC that were altered by the absence of Dp71 (Dalloz et al., 2001). To examine possible perturbations of the DAPC owing to the absence of Dp71 in mdx^3cv mice, we used specific antibodies against β-dystroglycan and α-syntrophin to analyze their respective localization and presence in spermatozoa. β-dystroglycan distribution was compared by indirect immunofluorescence, using the JAF polyclonal antibody and it was localized in the postacrosomal region and the flagellar middle piece of wild-type spermatozoa (Fig. 5A1). In contrast, β-dystroglycan immunostaining was observed as very faint fluorescence in the mdx^3cv spermatozoa (Fig. 5A2) and in some cells it was not detectable at all (data not shown), indicating low or no presence in the flagellar middle piece. Protein levels in mdx^3cv and wild-type spermatozoa were analyzed by western blotting. Only one protein was detected in the wild-type sperm membrane fraction as a band with of ∼50 kDa (Fig. 5D1), however, it decreased (Fig. 5D2) or was absent (data not shown) in membranes from mdx^3cv spermatozoa. As previously reported for guinea pig spermatozoa, β-dystroglycan is only present in sperm plasma membrane, therefore, it was not detected in spermatozoa without membranes from wild-type and mdx^3cv mice (data not shown).

In addition, the distribution of α-syntrophin was analyzed. Indirect immunofluorescence with C4 polyclonal antibody clearly revealed α-syntrophin staining concentrated in the middle piece of the flagellum and the postacrosomal region of whole sperm cells from the wild-type strain (Fig. 5B1), whereas in demembranated cells it was only localized in the flagellar middle piece (Fig. 5C1) as was observed for guinea pig spermatozoa (Hernández-González et al., 2001). The α-syntrophin immunoreactivity in whole spermatozoa from dystrophic animals was concentrated in the middle piece and the whole head (Fig. 5B2). In demembranated mdx^3cv spermatozoa α-syntrophin staining was relocalized in the postacrosomal region and along the flagella (Fig. 5C2). It was evident that α-syntrophin was clearly redistributed in mdx^3cv spermatozoa with normal or abnormal flagella (Fig. 5B2′). In addition, α-syntrophin staining was increased in the mutant strain (Fig. 5B2,C2), compared to the wild-type strain (Fig. 5B1,C1). To test whether in spermatozoa of the dystrophic strain, α-syntrophin was more concentrated than in the wild-type strain, they were analyzed by western blotting. Fig. 5 also shows the comparative level of α-syntrophin (55 kDa) in sperm membranes (E1,E2) and purified flagella (F1,F2) from wild-type and mutant mice, respectively. It is evident that syntrophin level was increased in both mdx^3cv α sperm fractions: membranes and flagella (Fig. 5E2,F2, respectively), compared with the same subcellular fractions from wild-type spermatozoa (Fig. 5E1,F1, respectively).

The results described above prove that the Dp71 mutation affects the DAPC level in the flagellar middle piece and show that α-syntrophin is abundant and redistributed along the flagellum and the head. These results indicate that the absence of Dp71 affects the DAPC components in opposite ways: decreasing the concentration of β-dystroglycan and increasing the concentration of α-syntrophin.

According to previous work Dp71, DAPs (β-dystroglycan and α-syntrophin), nNOS (Hernández-González et al., 2001), K⁺ channels (Félix et al., 2002) and AQP-7 (Calamita et al., 2001) are distributed within similar compartments of the spermatozoa plasma membrane. Moreover, in brain and muscle, α-syntrophin is thought to recruit several SAPs (Gee et al., 1998). It is known that the PDZ domain of syntrophins is involved in the interaction with these proteins (Brenman et al., 1996). As we found that α-syntrophin is clearly redistributed and more concentrated in mdx^3cv spermatozoa, we were interested in the effect of this alteration, or the absence of other DAPC components, on the presence and localization of different membrane ion channels and signaling proteins such as nNOS.

First, we investigated whether the absence of Dp71 modifies the distribution of voltage-dependent Na⁺ (μ1) and K⁺ (Kv1.1) channels and nNOS, three important SAPs involved in specific spermatozoa functions. Immunostaining of Na⁺ channels was observed in the postacrosomal region and the flagellar middle piece, and weak staining was present in the flagellar principal piece of wild-type spermatozoa (Fig. 6A1). However, in mdx^3cv spermatozoa, immunoreactivity to Na⁺ channels was concentrated in the whole head with less staining found in the middle piece (Fig. 6A2). The immunolocalization for K⁺ channels (anti-Kv1.1 antibody) in wild-type spermatozoa showed that it was concentrated in the whole head and slightly in the middle piece (Fig. 6B1). In mdx^3cv spermatozoa, K⁺ channel localization was completely altered, it was present in whole spermatozoa and specially concentrated in the acrosome region and in the middle piece (Fig. 6B2). It is remarkable that the redistribution of the Kv1.1 channels in spermatozoa of the mdx^3cv strain (Fig. 6B2) was similar to that observed for α-syntrophin redistribution (Fig. 5B2,C2). As Na⁺ and K⁺ channels are membrane proteins, they were not detected in demembranated spermatozoa (data not shown).

In the wild-type strain, nNOS immunostaining was detected in the flagellar middle piece and in the postacrosomal region (Fig. 6C1), as was Dp71 (see Fig. 3A1). In mdx^3cv spermatozoa, nNOS was more concentrated in the whole head and also in the middle piece (Fig. 6C2). In membrane-free spermatozoa of both strains, nNOS was immunolocalized in the middle piece, where mitochondria are condensed (data not shown). A clear redistribution and a considerable increase in staining were observed for these three proteins in mdx^3cv spermatozoa.

We analyzed the presence of K⁺ and Na⁺ channels and nNOS in membrane fractions of wild-type and mdx^3cv spermatozoa by western blotting and, as control of protein concentration, β-actin was detected simultaneously. Fig. 6 shows the detection of Na⁺ and K⁺ channels and nNOS in isolated membrane fractions from wild-type and mdx^3cv spermatozoa. It was evident that protein bands were expressed in membrane fractions, identified by their molecular weight: 210 kDa for the Na⁺ channels (Fig. 6D1,D2), 60 kDa for the K⁺ channels (E1,E2), 130 kDa for nNOS (F1,F2) and 45 kDa for β-actin (G1,G2). In membranes of mdx^3cv spermatozoa, Na⁺ channels and nNOS were clearly present in high concentrations, and K⁺ channels were increased to a lesser proportion. These results confirmed that the increased staining observed by immunofluorescence for Na⁺, K⁺ channels and nNOS was real. In addition, we provide the first detection and localization of a Na⁺ channel in mammalian spermatozoa.

Previous reports have shown that the absence of dystrophin gene products results in a loss of members of the DAPC from cell membrane in muscle, brain and retina (Blake et al., 2001). Our study aimed to compare the distribution of DAPC components in mdx^3cv spermatozoa, which lack the expression of all DMD gene products. We have shown that mouse spermatozoa exclusively contain dystrophin Dp71. To elucidate the role(s) of Dp71, we studied mdx^3cv spermatozoa and found that its absence resulted in: (1) an increase in abnormal flagellar morphology; (2) an increased level and redistribution of Up71, α-syntrophin, Na⁺ and K⁺ channels (μ1 and Kv1.1, respectively) and nNOS; and (3) a dramatic reduction in the level of β-dystroglycan.

The actin scaffold associated with the plasma membrane is important for sperm morphogenesis and differentiation (Ozaki-Kuroda et al., 2002), and its alteration produces defective sperm morphogenesis and male-specific infertility (Bouchard et al., 2000). Dp71 and β-dystroglycan are minor actin-binding proteins required for anchorage of the cytoskeleton to the plasma membrane through the DAPC (Howard et al., 1998; Chen et al., 2003). We showed that mdx^3cv mice produced an increased number of sperm with aberrant flagellar morphology, which could be a consequence of Dp71 absence and reduction of levels of β-dystroglycan. These alterations could modify the anchorage of the cytoskeleton to the plasma membrane. In agreement with this proposal is the rectification of the abnormal shape observed when the plasma membrane of mdx^3cv spermatozoa was removed, indicating the importance of the cytoskeletal interaction with the plasma membrane for flagellar morphology. Our results also show that the sperm alterations significantly reduced (almost 50%) the number of spermatozoa capable of fertilization. These abnormal spermatozoa had deficient motility, consisting of vibratory movements of the flagella without displacement (data not shown). This is the first description of a phenotype that results in a morphologic alteration of a non-muscular cell in dystrophic mice.

The most diminished or sometimes absent DAP in the mdx^3cv spermatozoa was β-dystroglycan, which agrees with a recent observation in retina, where it was severely reduced in Dp71 null mice (Dalloz et al., 2003). Conversely, the absence of Dp71 in mdx^3cv spermatozoa resulted in α-syntrophin relocalization and upregulation. It should be noted that the absence of Dp71 in mdx^3cv spermatozoa perturbs two members of the DAPC, β-dystroglycan and α-syntrophin, in different ways. In both cases, our results suggest that Dp71 is important for the formation and/or stabilization of the DAPC.

Members of the utrophin and dystrophin families are structurally homologous and it seems very likely that they perform similar functions. In mdx^3cv mice, the increased level of utrophin ameliorates the pathology of dystrophin deficiency (Tinsley et al., 1996). Up71, a member of the utrophin family, is relatively abundant in several non-muscular tissues including the testis (Wilson et al., 1999) and is able to bind DAPs such as β-dystroglycan, syntrophins and dystrobrevins (Blake et al., 2001). In mdx^3cv spermatozoa Up71 is upregulated and redistributed and may compensate for the absence of Dp71, forming a Up71∼DAPC with α-syntrophin, and in this way, Na⁺ and K⁺ channels and nNOS could be anchored to the plasma membrane from the head and middle piece. The similar Up71 and α-syntrophin relocalization in mdx^3cv spermatozoa allows us to speculate about dystrophin-like complexes organized to compensate for Dp71 absence (Fig. 7). Therefore, the K⁺ channel localized in the plasma membrane of the principal piece could only be anchored by α-syntrophin, because Up71 is absent in this sperm region (Fig. 7). These results suggest that in each plasma membrane compartment of mdx^3cv spermatozoa the Up71∼DAPC may be different, as described for the dystrobrevin complex in different regions of the kidney (Loh et al., 2001). Moreover, this complex may partially compensate for Dp71 absence in mdx^3cv spermatozoa with normal flagella and may be responsible for dystrophic mouse reproduction. However, Up71∼DAPC may also establish a different relationship with the sperm cytoskeleton, producing incorrect flagellar development. To prove this proposition, it is necessary to compare the in vitro fertilization capacity of both kinds of sperm cells, or of spermatozoa from mdx^3cv/utr^–/– mice.

Ion and water channels have a subcellular localization and tissue-specific expression patterns that determine their physiological roles, although, in spermatozoa, their localization and the mechanism of anchoring remain unknown (Darszon et al., 1999). In this study we found that Dp71∼DAPC may participate in the scaffold that contributes to anchor some membrane SAPs such as K⁺ and Na⁺ channels and nNOS in wild-type spermatozoa (see Fig. 7). Interestingly, all of them were upregulated in spermatozoa from mdx^3cv mice. These observations are different from those reported in retinal glial Müller cells of Dp71 null mice, where the K⁺ channel Kir 4.1 remained unchanged and aquaporin-4 was downregulated (Dalloz et al., 2003). The upregulation and redistribution of Na⁺ and K⁺ channels and nNOS may depend on a compensatory process for the Dp71∼DAPC alteration, which apparently does not correct the sperm morphologic phenotype.

The presence of voltage-dependent Na⁺ channels in mammalian spermatozoa has been suggested by the detection of sodium currents, which were inhibited by specific blockers (Shi and Ma, 1998; Patrat et al., 2000). The above evidence supports our result for the presence of a voltage-dependent Na⁺ channel (μ1), which was codistributed with Dp71∼DAPC in mouse sperm. To the best of our knowledge, it is the first time that a voltage-dependent Na⁺ channel was detected and localized by immunological procedures in mammalian sperm (Patrat et al., 2000). Its role in capacitation, acrosome reaction and/or motility remains to be determined.

The specific localization of ion channels suggests that they could produce ionic environments in specific sperm domains, which may be affected by their increased presence and redistribution, as shown here for spermatozoa from dystrophic mice. These altered ion environments may modify some sperm processes in mdx^3cv spermatozoa and consequently decrease their fertility. Hence, the adequate presence and localization of ion channels and the anchorage of the actin cytoskeleton seem very important for sperm physiology.

We are grateful to Gérald Hugon for antibody production, to Juan Carlos Garcia for mdx^3cv mouse care, to Aurora Candelaria for mice genotyping and to Ana Lilia Roa Espitia for her assistance in sperm preparation. Isabel Pérez Montfort corrected the English version of the manuscript. This work was partially supported by the CONACYT grant No. 34819 to E.O.H.-G. and The Association Francaise Contre les Myopathies to D.M. and A.R.