| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
Temple University School of Medicine, Department of Anatomy and Cell Biology, Philadelphia, Pennsylvania 19140
ABSTRACT
Previous studies of sperm from mice heterozygous for a t haplotype (t) and heterospecific combinations of the t complex identified two tightly linked genetic factors responsible for t/t male sterility related to expression of the flagellar waveform aberration, curlicue. Dnahc8, an axonemal dynein heavy chain gene, is a strong candidate for the proximal factor, Ccua, but the identity of the distal factor, Ccub, is unknown. In the present study, we employ motility assays of sperm from males heterozygous for t and novel heterospecific combinations of the t complex to demonstrate that Ccub is a composite of at least two synergic elements, Ccub1, positioned within a genomic interval spanning
0.6 Mb immediately distal to Dnahc8, and Ccub2, situated in a region 4–7 Mb distal to Ccub1. We also show that Tsga2, a testis-restricted gene, fulfills many of the prerequisites required to make it a strong candidate for Ccub1. These include: 1) its location within the aforementioned genomic interval; 2) a highly reduced level of testis expression by its heterospecific allele relative to the level of expression of its t allele; 3) determination that TSGA2t carries numerous nonsynonymous mutations in residues otherwise highly conserved in all known orthologous proteins; 4) the detection of major TSGA2 polypeptides in sperm protein extracts; and 5) the apparent distribution of these polypeptides in major sperm tail structures. Surprisingly, these TSGA2 isoforms appear to localize in the vicinity of the anterior acrosome, as well, suggesting that Tsga2 may also play a role in sperm-egg interaction. Finally, our results indicate that a TSGA2 polypeptide with apparent similarities to the smaller of the two sperm isoforms is expressed by epididymal cells.
epididymis, fertilization, sperm, sperm capacitation, sperm motility and transport
INTRODUCTION
Mouse t haplotypes (t) are naturally occurring structure/function variants of the t complex, an 40-Mb pair region at the proximal end of chromosome 17. Different t haplotypes, recently diverged from a common ancestor, share a structure consisting of four large, nonoverlapping inversions (In(17)1–In(17)4 from proximal to distal) relative to wild-type (+) homologs (Fig. 1). This common aberration in t haplotype chromosomal structure induces strong suppression of regional recombination in +/t heterozygotes with associative effects on the evolution of t complex genes [1].
|
Male t homozygotes are sterile because of the expression of inversion-bound factors that perturb numerous, distinct sperm functions, such as motility, zona pellucida binding, and penetration of the oolemma (Fig. 1) [2, 3]. The relatively recent discovery that particular heterospecific combinations of In(17)4 (products of recombination between Mus musculus wild-type chromatin and M. spretus wild-type chromatin) fail to complement t haplotype male fertility defects have made it possible to access the molecular basis of t-associated male germ cell traits formerly hidden from view by +/t recombination suppression [1]. In particular, studies of the motility of sperm from males heterozygous for heterospecific recombinant In(17)4 homologs and a t haplotype have permitted high-resolution mapping of genetic factors leading to "stop," an abnormality in oolemma penetration exhibited by sperm from t/t males, and curlicue, a Ca2+-sensitive sperm tail curvature aberration frequently considered the signature phenotypic reflection of t/t male infertility [4–8].
Current theory states that at least two factors in the Hybrid Sterility 6 (Hst6) locus conspire to cause curlicue [5, 6]. A strong candidate for the more proximal of the two factors (known as Curlicue a, abbreviated Ccua) is Dnahc8, a gene coding for an axonemal outer arm 
dynein heavy chain exhibiting testis-restricted expression and confinement of its major protein isoform to the principal piece of the sperm tail [6–10]; however, the distal factor, Curlicue b (Ccub), has remained refractory to identification.
To identify Ccub candidates, we have studied the motility of sperm from males heterozygous for t and novel heterospecific recombinant In(17)4 homologs. The results of these experiments indicate that Ccub has a more complex molecular basis than previously contemplated, apparently consisting of at least two synergic elements. The more proximal of these (Ccub1) maps just distal to Dnahc8 in Hst6, and the more distal element (Ccub2) localizes outside of Hst6 in the considerably more distal Hst5 locus [5]. The genomic interval encompassing Ccub1 houses Tsga2 (Fig. 1), a gene whose protein products were believed to be confined to the testis in male mice [11, 12], but whose orthologous proteins from Cyprinus carpio and Ciona intestinalis were shown to localize to the sperm tails of those organisms [12, 13]. Interestingly, the genomic interval surrounding Ccub1 is coincident with the region circumscribing the Stop1p factor, an element of the stop phenotype, whose t allele is thought to interact synergetically with the t alleles of other genes to abrogate the ability of sperm from t/t males to penetrate zona pellucida-free mouse eggs in vitro [3, 6].
To determine whether Tsga2t plays a central role in the manifestation of curlicue or other t haplotype-associated sterility phenotypes, we undertook an in-depth study of its sequence and expression in the mouse. Our results not only indicated important sequence differences between Tsga2t/TSGA2t and all other known alleles/orthologous proteins, and a haploinsufficient pattern of expression of its M. spretus allele in the testes of males heterozygous for that allele and a t haplotype, but demonstrated the presence of two major TSGA2 isoforms in the mouse sperm tail, as well as in the sperm head in the vicinity of the anterior acrosome. Interestingly, this study also indicated that a novel TSGA2 isoform with properties similar if not identical to the smaller sperm isoform is expressed by epididymal cells.
MATERIALS AND METHODS
Mice, Genotyping, Assessment of Flagellar Curvature and Progressive Velocity, and Statistical Analysis
Mice used in this study were raised and maintained in the colony of S. Pilder at Temple University School of Medicine in accordance with NIH regulations/guidelines and IACUC protocols. Although the complete t haplotypes, tw5 and tw32, the wild-type chromosome 17 homolog, and the heterospecific chromosome 17 homologs SR2, SR3, SR5 (previously called SR4), SR8, and SR14 are presently maintained as congenics on the C57BL/6 genetic background, new sperm motility analyses of all heterozygous genotypes were performed on animals at the N2-N6 backcross generations to C57BL/6. This is in keeping with tests of sperm from males of the SR12 line, surviving only to the N6 backcross generation to C57BL/6. All heterozygous genotypes tested carried the t haplotype, tw32.
All experimental animals were genotyped by restriction fragment length polymorphism analysis of Southern blotted mouse tail-tip DNA or PCR using a variety of previously described probes that can distinguish between M. musculus wild-type, t haplotype, and M. spretus DNA at informative marker loci [5–10, 14].
To assess flagellar curvature, the percentage of curlicue sperm tails from each heterospecific recombinant chromosome 17/t heterozygous genotype was measured as in Olds-Clarke and Johnson [4] and Redkar et al. [6]. Briefly, 20 min after release of sperm from the cauda epididymides (before full capacitation) into 400 ml of Krebs-Ringer bicarbonate medium containing 1.7 mM Ca2+ and 2% BSA at 37°C in 5% CO2 in air, an aliquot of each sperm suspension was diluted and videotaped for 5 min at a magnification of x256. Each taping segment was coded, and a trained observer not involved in the videotaping viewed the coded tape segments and classified each motile sperm as having either a straight or a curlicue flagellum. Sperm aliquots from 5 to 10 animals of each genotype were tested. The mean percentage of curlicue flagella ± the SEM was calculated for each genotype. Significant differences between the means for each genotype were determined by the Tukey honestly significant difference (HSD) test after ANOVA.
To assess progressive motility, the mean straight-line velocity (VSL) was determined for sperm from each male with a Hamilton-Thorn Motility Analyzer as in Olds-Clarke and Johnson [4]. Statistical differences between the means for different genotypes were determined by the Tukey HSD test after ANOVA.
RNA Methods and Sequencing
Differential display-PCR experiments were performed as described in Fossella et al. [8], except the RNAs used were from the testes of C57BL/6-+/+ and C57BL/6-SR3/SR3 males. Northern blot, differential display-PCR, RT-PCR, and 5'-RACE experiments were performed as previously described [8, 9]. Briefly, total RNAs were extracted from mouse testicular or epididymal tissue from 8- to 10-wk-old mice by the method of Chomczynski and Sacchi [15], treated with RNase-free DNase (Promega) at 37°C for 30 min, and either separated in 1.5% agarose-formaldehyde gels and blotted to positively charged nylon membranes (Amersham), or subjected to RT-PCR and RACE. For the Tsga2 probe used in Northern analysis and all RT-PCR experiments, the region between the 73rd and 1036th base of the published testicular Tsga2 mRNA sequence (containing the entire Tsga2 coding region; RefSeq Accession No. NM_025290) was amplified using the following primer set: Tsga2CF: 5'-gctcctgagatcctgtctgaaacc-3'; Tsga2CR: 5'-tgtgtctgtggtcttctgcctc-3'. To obtain the full length testicular Tsga2 transcripts from various strains and species of mice, Tsga2 cDNAs were prepared from total testis or epididymal RNAs with either the SMART RACE cDNA Amplification Kit (Clontech) or the Gene Racer Kit (Invitrogen), according to the manufacturer's instructions. First strand testis cDNAs from BALB/c aqnd SW mice were purchased commercially (Origene; Clontech). Amplified mRNAs were inserted into a Zero Blunt TOPO vector (Invitrogen) and sequenced as previously described [9].
TSGA2 Antisera Preparation and Test for Antisera Specificity
Two monospecific polyclonal antisera (
N and C) were raised in rabbits against KLH-linked peptides (Invitrogen) from regions near the N- and C-termini, respectively, of mouse testicular TSGA2 (RefSeq Accession No. NP_079566). Each peptide was selected on the basis of its computationally determined antigenicity, its location in the polypeptide, and its potential lack of cross-reactivity with other mouse proteins. The positions and sequences of the peptides are: for N, residues 16–30, DLGEYEGERNEVGER; and for C, residues 275–289, RQESQENSYDIDQGN. Following ELISA, each antiserum was affinity-purified on and eluted from a column of Sepharose-4B coupled to the peptide against which it was raised and subsequent OD280 reading. Specificity of each antisera was further verified by preadsorption of each antisera with its respective peptide as in Kim et al. [16] before use in Western blot analysis.
Testis, Epididymis, and Sperm Protein Extract Preparation, Electrophoresis, and Western Blot Analysis
Freshly isolated testes (decapsulated) or epididymides (complete or sperm depleted) were Dounce homogenized on ice in RIPA-protease/phosphatase inhibitor buffer without SDS (50 mM Tris-HCl pH7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% NP-40, 1% Na-deoxycholate, 1 mM PMSF, 5 µg/ml Aprotinin, 5 µg/ml Leupeptin, 5 µg/ml Pepstatin, 1 mM Na3VO4, 1 mM NaF). Homogenates were kept at 4°C for 2 h with frequent vortexing, then centrifuged. After determination of supernatant protein concentrations and the addition of fresh protease/phosphatase inhibitors, aliquots (30–100 µg) were further prepared for gel electrophoresis. For sperm, freshly isolated cauda epididymides were lightly minced, then incubated at 37°C in IVF medium (modified Krebs-Ringer-bicarbonate buffer containing 2% BSA, 1.7 mM Ca2+, and 50 mM Hepes) [4] in 5% CO2 in air for 20 min to allow sperm to swim out. The sperm solutions were cleared of debris by brief, low-speed centrifugation, the supernatants containing sperm were collected, and sperm were counted. Supernatants were then centrifuged at 16000 x g and sperm pellets were resuspended in RIPA-protease/phosphatase inhibitor buffer without SDS plus protease inhibitor cocktail at a concentration of 7.7 x 104 sperm/µl. Aliquots containing 1 x 106 sperm were further prepared for gel electrophoresis.
For sperm fractionation experiments, we used the Compartmental Protein Extraction Kit (Chemicon International Inc.), a subcellular fractionation technique that utilizes a series of proprietary buffers to separate (enrich) cytoplasmic (soluble), nuclear (soluble), membrane, and cytoskeletal plus nuclear matrix protein fractions from each other based on their differential solubilities in particular salts and/or detergents. After fractionation following the manufacturer's instructions, the postcytoskeletal pellet was boiled in 2% SDS with reducing agent and centrifuged briefly, and the resultant supernatant was designated the highly insoluble "cytoskeletal 2" fraction. The protein concentration of each fraction was measured, and aliquots of each were stored at –80°C before electrophoresis.
The results of the Compartmental Protein Extraction Kit procedure were verified with the differential centrifugation technique of Visconti et al. [17]. Briefly, cauda epididymal sperm pellets were washed in a medium containing 20 mM Tris-HCl, 130 mM NaCl, 1 mM EDTA, pH 7.5, centrifuged briefly, resuspended in 20 mM Tris-HCl and 1 mM EDTA, and homogenized on ice by 10 strokes of a Teflon/glass homogenizer. After homogenization, samples were centrifuged at 10000 x g for 10 min at 4°C and the pellets were designated P10. The supernatants of the 10000 x g spin were further centrifuged for 1 h at 100000 x g at 4°C. Both the top half of the supernatant fraction and the pellet fraction were saved and were designated S100 and P100, respectively. The P10 fraction corresponded to large proteins and insoluble protein complexes found in the sperm cytoskeleton, nuclei, mitochondria, and sperm that survived homogenization. The S100 fraction contained primarily soluble (designated cytoplasmic) proteins, and the P100 contained primarily membrane proteins.
Fibrous sheath (FS) and outer dense fiber (ODF) fractions were isolated following the protocol developed by Oko [18]. Briefly, the heads and tails of 1 x 109 sperm released from the caudae of 10 mice were separated by sonication at 100% duty on ice with four 15-sec bursts, each burst separated from the next by a 30-sec interval. After low-speed centrifugation, the supernatant containing sperm heads was collected separately from the pellet, which contained primarily tails. Two rounds of centrifugation of the pellet through 65%–75% sucrose step gradients at 100000 x g for 70 min yielded an interphase fraction containing nearly 100% sperm tails, as determined by phase contrast microscopy. The tail fraction was pelleted in two tubes, and one pellet was resuspended in 2% Triton X-100, 5 mM dithiothreitol (DTT), 50 mM Tris-HCl, pH 9.0 plus protease/phosphatase inhibitors, mixed at 4°C for 15 min, then centrifuged briefly. After this step was repeated, the pellet was resuspended in 50 mM Tris-HCl, pH 9.0, and centrifuged. The pellet was resuspended and mixed at 4°C for 4 h in 4.5 M urea, 5 mM DTT, 25 mM Tris-HCl, pH 8.0, plus protease inhibitors. After centrifugation at 16100 x g for 5 min, the pellet was resuspended in 25 mM Tris-HCl, pH 8.0, layered on a 35%–75% sucrose step gradient, and centrifuged at 100000 x g for 1 h. The interface, highly enriched for FS proteins, was isolated, diluted in 25 mM Tris-HCl, pH 8.0, and pelleted at 50000 x g for 10 min. The FS pellet was saved at –80°C for electrophoresis. To prepare the ODF fraction, the second half of the tail pellet was mixed at RT for 1 h in 50 µl of 1% SDS, 2 mM DTT, 25 mM Tris-HCl, pH 8.0, plus a cocktail of protease inhibitors, followed by centrifugation at 16100 x g for 5 min. The pellet was resuspended in 25 mM Tris-HCl, pH 8.0, layered on a 35%–75% sucrose gradient, and treated from there identically to the FS preparation.
For electrophoresis, all aliquots were resuspended in 1x SDS-PAGE sample buffer (Invitrogen) either with or without reducing agent (Invitrogen) and boiled. After boiling and brief centrifugation, reducing agent was added to the supernatants not containing it, and preformed 10% NuPAGE Bis-Tris polyacrylamide gels (Invitrogen), mounted in an Xcell SureLock apparatus (Invitrogen), were immediately loaded. Proteins were separated at 150 V constant for 90 min with 1x SDS running buffer (Invitrogen). Molecular weight standards were purchased from Invitrogen and Amersham.
Proteins separated by electrophoresis were transferred to Immobilon-P PVDF membranes (Millipore) in 1x transfer buffer (Invitrogen) at 4°C, 30 V constant, for 1 h in the same apparatus reset for Western blotting. Membranes were then soaked briefly in methanol, dried thoroughly, wet in 20% methanol, then stained in Ponceau-S/5% acetic acid to detect total protein. Stained membranes were blocked in 5% modified dried milk (Genotech) in TBS-T buffer (10 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% Tween-20) before being probed with an affinity-purified primary TSGA2 or control antiserum followed by a peroxidase-conjugated secondary antibody. After washing, the signal was visualized using an enhanced chemiluminescent system (Pierce) followed by exposure to CL-Xposure film (Pierce). Band intensities on films were measured by scanning densitometry.
Antiserum against human ß-tubulin (H-235) used as a positive control antibody for gel loading and a negative control for the purity of FS and ODF fractions was purchased from Santa Cruz Biotechnology. Antisera against mouse A kinase anchoing protein 4 (AKAP4; anti-AKAP82) [19] and ODF2 (anti-450) [20], used as positive controls for the purity of FS and ODF fractions, respectively, and as markers of the detergent solubility of proteins in the two cytoskeletal fractions produced in the modified compartmental fractionation procedure were the kind gifts of Drs. George L. Gerton, Center for Research on Reproduction and Women's Health, University of Pennsylvania Medical Center, Philadelphia, PA, and Frans A. van der Hoorn, Department of Biochemistry and Molecular Biology, University of Calgary, Calgary, Alberta, Canada, respectively.
Fluorescent Immunocytochemistry
For immunocytochemistry, cauda epididymal sperm in IVF medium were centrifuged at 800 x g, and the sperm pellets were resuspended in 200 µl PBS. After counting, the sperm concentration was adjusted to 2 x 107/ml, and aliquots of 50 µl were pipetted into small circles drawn with a hydrophobic barrier pen on polylysine-coated slides. These were then incubated in humid chambers for 10 min at room temperature. To extract mitochondria, half the slides were treated with a high pH solution (200 mM Tris HCl pH 9.5, 2 mM DTT, 0.1% Triton X-100) [21, 22] in a humid chamber for 10 min at room temperature. Following the removal of excess liquid, high pH-treated and untreated sperm were fixed in freshly made 4% paraformaldehyde in PBS at 4°C for 1–2 h. Fixed, air-dried sperm were then immersed in chilled acetone for 10 min, redried, then washed in two changes of 0.1% Tween-20 in PBS and one change of PBS. Antigens were then retrieved by boiling the slides in 10 mM citrate buffer pH 6.0 for 10 min, cooling, and washing several times in 0.1% Triton X-100 in PBS followed by blocking in a solution of 5% BSA, 10% goat serum in PBS for 4 h at room temperature (sperm tested for TSGA2 external surface antigenicity were neither permeated with acetone, washed in Triton X-100 postantigen retrieval, nor depleted of mitochondria). To account for possible loss of antigenicity caused by postantigen retrieval Triton X-100 washes, fluorescence experiments were repeated without the Triton X-100 washes using sperm with intact mitochondria. Following removal of the blocking solution, either primary antibody or preimmune serum diluted in blocking solution was added, and slides were incubated at 4°C overnight in humid chambers. The next day slides were washed three times in 0.1% Tween-20 in PBS, after which a 1:500 dilution of Cy3-labeled goat anti-rabbit secondary antibody (Jackson ImmunoResearch) in 1% BSA, 5% goat serum in PBS was added to each slide. Slides were then incubated at room temperature for 3 h in the dark, followed by 10 washes in ice-cold PBS for 10 min each wash. Slides were then mounted in an antiquenching mounting solution (Gel Mount; Biomeda) containing DAPI for fluorescence microscopy with a Nikon Eclipse E800.
Computational Methods
To identify non-TSGA2 mouse proteins that may bind to TSGA2 antisera raised against specific peptides, the BLAST algorithms [23] TBLASTn and BLASTp were used with the peptides DLGEYEGERNEVGER and RQESQENSYDIDQGN serving as query sequences. Analyses employed the NCBI mouse genome sequence build 34 and the nonredundant protein database as subjects, a search for near-exact matches, and an E value = 10. To identify TSGA2 orthologs, a similar variety of BLAST algorithms was used (Psi-BLAST, Gapped-BLAST, and BLASTp) with M. musculus TSGA2 (RefSeq Accession No. NP_079566) serving as query sequence with an E value = 0.1. BLAST analyses were performed at the NCBI website (www.ncbi.nlm.nih.gov) and at the Ensembl Genome Browser website (http://www.ensembl.org). Some orthologous sequences were abstracted directly from the literature [12]. In addition to the NP_079566 sequence for M. musculus, the M. musculus strain-specific 129S1, Balb/C, C57BL/6 and Swiss Webster (SW) wild-type alleles, the t haplotype alleles tw32 and tw5, and an outbred M. spretus wild-type allele were isolated as described above under RNA Methods and Sequencing (GenBank Accession Nos. DQ141549, DQ141550, DQ141555, DQ141551, DQ141553, DQ141554, and DQ141552, respectively). Allelic and orthologous polypeptides were aligned by CLUSTAL W (http://bioweb.pasteur.fr/seqanal/interfaces/clustalw.html) [24]. Domains/motifs in TSGA2 were identified with the ELM program [25] available at the EXPASY website (http://us.expasy.org).
RESULTS
Ccub Activity is Dependent on the Expression of More than One Genetic Element
In previous studies [6, 7], we suggested that the curlicue phenotype required the expression of at least two tightly linked t haplotype factors in the Hst6 locus, one, Ccua, proximal to D17Mit135 at 29.70 Mb, and the other, Ccub, proximal to D17Mit191 at 30.09 Mb but distal to D17Mit146 at 29.92 Mb (Fig. 2A). To confirm the map position of Ccub, we derived additional heterospecific recombinant chromosome 17 homologs (SR8, SR12, and SR14), and tested males heterozygous for each new recombinant and the t haplotype, tw32, for the percentage of sperm released from the cauda epididymides expressing the curlicue phenotype (Fig. 2B).
|
SR8 [6] was derived from SR3, carrying the SR3 proximal M. spretus end, but a different M. spretus distal breakpoint between D17Mit63 (31.11 Mb) and D17Mit16 (32.09 Mb) (Fig. 2B). Thus, this recombinant homolog carried both Ccua and Ccub as defined in Redkar et al. [6]. Surprisingly, sperm from males of the SR8/tw32 genotype exhibited high curlicue levels, not significantly different from levels displayed by sperm from SR2/tw32 and SR5/tw32 males, but significantly lower than levels exhibited by sperm from SR3/tw32 (Fig. 2B) and tw5/tw32 males [4]. Taken together with our earlier findings [6, 7] as summarized in Figure 2A, this result indicated that either Ccub was located distal to the SR8 M. spretus distal breakpoint in the Hst5 locus, contrary to our previously reported estimate of its position, or it consisted of more than one element.
To distinguish between these two possibilities, we examined sperm from males heterozygous for either SR12 or SR14 and tw32 for expression of the curlicue phenotype (Fig. 2B). Both SR12 and SR14 were derived from SR2, the former carrying an extensive proximal portion of SR2 including its proximal M. spretus end, and the latter carrying a small distal portion of SR2 including its distal M. spretus end (Fig. 2B). In addition, the distal M. spretus end of the former and the proximal M. spretus end of the latter overlapped at Pou5F1 (34.02 Mb; formerly Oct-4), but not at D17Mit233 (34.46 Mb), where SR12 was M. musculus wild-type and SR14 was M. spretus (data not shown). Interestingly, sperm from both SR12/tw32 and SR14/tw32 males expressed low percentages of curlicue flagella, not significantly different from sperm of males of the +/tw32 genotype for this characteristic t/t phenotype, but significantly lower than sperm of males of the SR5/tw32, SR2/tw32, SR8/tw32, and SR3/tw32 genotypes (Fig. 2B). In addition, sperm from males of the SR3/tw32 genotype exhibited a significantly higher percentage of curlicue flagella than sperm from males of the SR5/tw32, SR2/tw32, or SR8/tw32 genotypes, the latter three not being significantly different from each other (Fig. 2B). These findings were similar to those previously reported [6] (summarized in Fig. 2A), thus assuring us that the moderately high percentage of curlicue flagella previously observed for sperm from SR5/tw32 and SR2/tw32 (C57BL/6 x 129S1) F1 males and the significantly higher percentage of curlicue flagella observed for sperm from SR3/tw32 (C57BL/6 x 129S1) F1 males and tw5/tw32 (C57BL/6 x 129S1) F1 males were not artifacts of the F1 background [6]. In addition, these data, in concert with the previously reported findings summarized in Figure 2A [4, 6], supported the idea that Ccub function depended on the synergic activity of at least two elements, one proximal (Ccub1), mapping between the distal end of Dnahc8 at 29.48 Mb and D17Mit191 at 30.09 Mb, and one distal (Ccub2), mapping to SR14 between 34.02 Mb (Pou5F1) and 37.50 Mb (the approximate distal end of the t complex) [26]. Interestingly, the Ccub1 element mapped to the same genomic interval as Stop1p, a factor of unknown molecular identity thought to be complicit in the expression of the stop sperm-oolemma penetration abnormality exhibited by sperm from t/t males.
To test whether SR12/tw32 and SR14/tw32 males produced sperm that exhibit subtle motility defects, the mean progressive velocities (mean VSL ± SEM, m/sec) of sperm from males of these genotypes were compared to the mean VSLs of sperm from males of the SR5/tw32, SR2/tw32, SR8/tw32, SR3/tw32, and +/tw32 genotypes. Our results showed that the mean VSLs for sperm from males of the SR14/tw32, SR5/tw32, SR2/tw32, SR8/tw32 and SR3/tw32 genotypes were not significantly, different from each other (32.5 ± 0.9, 32.8 ± 2.2, 27.9 ± 1.8, 30.6 ± 1.3, and 25.5 ± 1.3 m/sec, respectively), but were significantly lower than the mean VSLs of sperm from males of both the +/tw32 and SR12/tw32 genotypes (56.26 ± 3.2 and 47.7 ± 1.7 m/sec, respectively). Additionally, the mean VSL of sperm from males of the +/tw32 genotype was slightly, but significantly, higher than the mean VSL of sperm from males of the SR12/tw32 genotype. These findings suggested that although sperm from both SR12/tw32 and SR14/tw32 males exhibited negative control levels of curlicue flagella, they did exhibit other, more subtle motility defects.
Interestingly, additional breeding studies demonstrated that although SR2/t males were sterile, SR2/SR2 males were fertile, producing levels of curlicue sperm not significantly different from fertile +/t negative control levels (data not shown). When these results were considered together with those of the aforementioned motility studies shown in Figure 2B, a model describing the molecular basis of Ccub function became apparent: not only would Ccub consist of at least two synergic elements mapping to the proscribed genetic intervals, but strong candidates for Ccub elements would have to demonstrate several additional properties: 1) the heterospecific (M. spretus) alleles of these synergic elements would have to exhibit haploinsufficient, but not null, expression in the M. musculus genetic background; 2) the protein products expressed by its t alleles should demonstrate nonsynonymous mutations at sites that are nearly completely conserved in orthologous proteins, thus leading to defective function; and 3) the protein products of Ccub genes should demonstrate, at least in part, sperm tail localization.
Low Steady-State Levels of Tsga2 Testis mRNA and Protein are Produced in Males Homozygous for Heterospecific Recombinant In(17)4 Homologs Containing the M. spretus Allele of Ccub1
Because our model of Ccub function specified that the M. spretus alleles of its synergic elements must be poorly expressed relative to their t allelic counterparts in the laboratory mouse genetic background, we performed differential display-PCR with testis RNAs from SR3 vs. wild-type homozygotes (data not shown) to attempt to isolate Ccub candidate genes. Of several differentially expressed genes isolated, the only one mapping to the genomic interval containing Ccub1 (Fig. 2B) was Tsga2, a gene whose protein expression had been reported to be testis-restricted [11]. We verified the differential expression of Tsga2S vs. Tsga2t by Northern and Western blot analyses (Fig. 3). In all lines tested, those homozygous for the M. spretus allele of Tsga2 (regardless of the allelic nature of Ccub2 and Dnahc8) expressed at best 10% of the steady state levels of either Tsga2 mRNA or TSGA2 protein produced by the testes of either their wild-type or t homozygous counterparts. Thus, the poor but nonnull expression of the M. spretus allele of Tsga2 in the C57BL/6 genetic background satisfied one of the three conditions required by our model of Ccub function for strong Ccub1 candidacy.
|
The tw5 and tw32 Isoforms of TSGA2 Differ Significantly from All Other Known Orthologous Proteins in Two Acid-Rich Regions Making Up the Flanks of the Highly Conserved N-Terminal Two-Thirds of the Protein
To test Tsga2 for its ability to satisfy the second condition of our model, we first isolated and sequenced Tsga2 cDNAs from C57BL/6-+/+, C57BL/6-+/tw32, and C57BL/6-+/tw5 congenic mouse testes. The putative translation products of tw32 and tw5 cDNAs contained 15 and 13 amino acid differences, respectively, relative to the wild-type product (Fig. 4). All of the tw5 mutations were identical to the first 13 alterations in the tw32 isoform, and these were contained within the N-terminal two thirds of the protein, a region comprised of seven tandem Membrane Occupation Recognition Nexus (MORN) repeats flanked by two short, acid-rich regions (ARRs), previously shown to be highly conserved among TSGA2 orthologous proteins [12].
|
To further assess the possible significance of any of the N-terminal 13 mutations in t alleles, we isolated and sequenced Tsga2 cDNAs from the testes of the M. musculus strains 129S1-+/+, C57BL/6-SR2/SR2, BALB/c-+/+, and SW-+/+, and aligned their putative translation products with those from C57BL/6-+/+, tw32, tw5, and known orthologs from other species including fish, rat, human, chimpanzee, dog, chicken, and sea squirt (Fig. 5). Three t mutations, two in the proximal and one in the distal MORN motif-flanking ARRs, altered otherwise completely-conserved acidic amino acids to a tyrosine, a glycine, and a lysine, respectively. Two other t mutations, both in the distal ARR, altered the amino acids arginine and glutamate, completely conserved in all wild-type mammalian sequences, to cysteine and alanine, respectively. In further support of our model, the M. spretus TSGA2 isoform isolated from SR2 was wild-type at all of those residues. The nearly complete conservation of the wild-type ARR residues altered in TSGA2t isoforms to nonsynonymous mutations argues that the wild-type residues were indispensable to normal TSGA2 behavior.
|
At Least Two TSGA2 Isoforms Are Present in Sperm Released from the Cauda Epididymides of +/+ and tw5/tw32 Mice
Two previous studies have suggested that TSGA2 is confined to testicular germ cells, i.e., it is not present in mature sperm of mice [11, 12]. However, the second of these studies and the work of Satouh et al. [13] showed that fish and sea squirt TSGA2 orthologous proteins, respectively, were located in the sperm tails of these animals, the latter study also indicating that sea squirt TSGA2 was probably a component of the axonemal radial spoke complex. Nonetheless, none of these studies tested for the presence of TSGA2 in mouse sperm protein lysates by Western blot analysis with antibodies raised against mouse TSGA2. Therefore, we probed Western blots of both testis and sperm total protein lysates from +/+ and tw5/tw32 mice with affinity-purified antibodies raised against two highly antigenic peptides (see Materials and Methods), one from each of the N- and C-terminal regions (N and C, respectively) of mouse testicular TSGA2+ (Fig. 6A). In testis extracts from mice of both genotypes, N recognized two polypeptide bands running very close to each other at 40 kDa, similar in size to the previously identified polypeptide(s) [11, 12], whereas C identified only the slightly larger and more abundant of the two. These findings suggested that the somewhat smaller polypeptide was truncated at its C-terminus. The protein identified by both antibodies migrated more slowly by SDS-PAGE (by 5–6 kDa) than was predicted by computational analysis of the TSGA2 sequence, a feature of testicular TSGA2 that had been reported previously [11]. When we probed Western blots of sperm protein extracts, we obtained a surprisingly different result: N recognized only the slightly larger of the two testicular bands, whereas C identified not only this same band, but an entirely novel, 3–4-fold more abundant smaller polypeptide of 30 kDa (Fig. 6A).
|
To test for the possibility that our antisera were binding to peptide epitopes other than those against which they had been raised, each antisera was preadsorbed with a 100-fold molar excess of its respective peptide [16], and the Western blots were repeated (Fig. 6B). The results demonstrated that the antisera bound to proteins carrying only the peptides against which the antisera had been raised. Preadsorption of C with the N peptide and vice versa did not prevent either antiserum from binding to the same pattern of proteins shown in Figure 6A, left panel (data not shown).
To ascertain whether our antisera might be binding to non-TSGA2 mouse proteins containing peptide epitopes similar to those against which our antisera were raised, the N and C peptide sequences were used to query the mouse genome sequence (build 34) and the nonredundant protein database by TBLASTn and BLASTp analyses, respectively (see Materials and Methods). Based on the criterion that at least five contiguous amino acids in a protein must be identical to any five contiguous amino acids of either the N or C peptide for subject proteins to have sufficient antigenicity to bind to our antisera, our analyses uncovered a total of five candidate mouse proteins other than TSGA2. Of these five, none were close to either 30 or 40 kDa in molecular mass (instead ranging from 54 to 500 kDa) and none were known to be expressed in the testis (data not shown). Given that neither the mouse genome sequence nor the nonredundant protein database is complete, these assays cannot be considered entirely exhaustive. Nonetheless, the results do imply that the polypeptides binding to our anti-TSGA2 antisera were TSGA2 isoforms.
Major Sperm Isoforms of TSGA2 are Distributed in Both the Flagellum and the Head in the Vicinity of the Anterior Acrosome
To determine the subcellular location of TSGA2 polypeptides in sperm, and to demonstrate that they were neither testicular nor epididymal cell contaminants of the sperm lysates, sperm from males of both +/+ and tw5/tw32 genotypes were collected from cauda epididymides by swim out, then fixed in formaldehyde, permeated with acetone, washed in Triton X-100 after antigen retrieval, and subjected to immunofluorescence (see Materials and Methods). For N, sperm from males of both genotypes exhibited bright, intense fluorescence over the proximal principal piece of the tail with little if any fluorescence over the midpiece (Fig. 7A). No significant difference in fluorescence was apparent whether mitochondria were extracted before fixation or not (data not shown), and no appreciable fluorescence was evident when N-preimmune serum was used (Fig. 7E). For C and its preimmune serum counterpart, the results were similar to those for N and its counterpart, except that C fluorescence was strong over both the proximal principal piece and the midpiece, although slightly less intense in the midpiece (Fig. 7, C and G, respectively). When N and C were used to probe sperm from which the mitochondria had not been extracted and which were neither permeated with acetone nor washed with Triton X-100 postantigen retrieval, no appreciable fluoresence was evident (Fig. 7, I and K, respectively).
|
The results of these initial indirect immunofluorescence studies implied the following: 1) the polypeptide bands observed on Western blots of sperm extracts were not products of contaminating epididymal or testicular cells; 2) the 40-kDa TSGA2 isoform and at least a portion of the 30-kDa isoform localized to different lengthwise compartments in the flagellum; and 3) the peptide epitopes recognized by N and C were not exposed on the sperm surface.
Because Triton X-100 is capable of solubilizing membranes, thus resulting in a loss of antigenicity by virtue of removing proteins closely associated with the solublized membranes, the fluorescence experiments were repeated without employing postantigen-retrieval Triton X-100 washes on sperm containing intact mitochondrial sheaths. For N, sperm from males of both genotypes exhibited bright, intense fluorescence over the proximal principal piece of the tail, as was the case when Triton X-100 washes had been employed. However, faint but clearly detectable fluorescence was visualized throughout the midpiece and in the vicinity of the anterior acrosome (Fig. 7, M, N, and O). No appreciable fluorescence was evident when N-preimmune serum was used (data not shown). For C, our new results were even more striking. Not only was fluorescence intensified over the entire midpiece relative to experiments using Triton X-100 washes, but strong fluorescence was evident in the connecting piece and over the anterior acrosome (Fig. 7, P, Q, and R). Again, no appreciable fluorescence was evident when C-preimmune serum was used (data not shown). Taken together, these results implied that a minor portion of the larger isoform and a much more substantial portion of the smaller isoform might be present in the region bounded by the cytoplasmic aspect of the plasma membrane and the external aspect of the outer mitochondrial/acrosomal membranes, or perhaps, for the smaller isoform, within the mitochondrial sheath. Another possibility was that either or both isoforms could be integral to the plasma and/or outer mitochondrial membranes, without antigenic epitopes exposed on their outer surfaces.
To gain greater insight into the location of each of the TSGA2 isoforms in sperm, we partitioned total sperm proteins into subcellular fractions by two methodologically different but complementary procedures, one based on the differential solubilities of proteins in buffers containing various concentrations of salts and/or detergents (compartmental fractionation), the other involving fractionation of cellular components by differential pelleting in hypotonic buffer (differential centrifugation) (Fig. 8A). Because the concentrations of each component in the extraction buffers employed in compartmental fractionation were proprietary and thus unknown, and because many cytoskeletal components of mammalian sperm are known to be difficult to solubilize, we added a second cytoskeletal extraction to the compartmental fractionation procedure, designed to solubilize components of the sperm cytoskeleton that were highly resistant to detergent solubilization (i.e., still in the pellet from the first cytoskeletal extraction; see Materials and Methods). We also tested both cytoskeletal fractions for the presence and abundance of sperm tail cytoskeletal proteins with antibodies raised against known components of the ODF (ODF2) and the FS (AKAP4) (see Materials and Methods). By both procedures, the 40-kDa protein fractionated nearly exclusively into the cytoskeletal fraction (the first cytoskeletal fraction by compartmental fractionation), whereas the 30-kDa isoform partitioned with the cytoplasmic (soluble) fraction by differential centrifugation and with both the cytoplasmic (soluble) and the second cytoskeletal fractions by compartmental fractionation. Both ODF2 and AKAP4 partitioned primarily into the second cytoskeletal fraction (with a minor presence of each in the first cytoskeletal fraction) by the compartmental fractionation procedure, attesting to their highly insoluble natures (Fig. 8A). These results suggested that the majority of both TSGA2 isoforms could be elements of discrete cytoskeletal structures in sperm or differentially soluble components of the same structures. In addition, these data all but eliminated the possibility that either isoform was an integral membrane protein or a nuclear component, although one or both isoforms could exist, in part, in complexes with integral/peripheral membrane proteins.
|
In addition, N recognized an 37-kDa polypeptide in the fractions enriched for cytoskeletal proteins by both fractionation procedures (first cytoskeletal fraction by compartmental fractionation). A polypeptide band of the same size was previously observed on Western blots of total sperm protein extracts only after relatively long exposures, and then in extraordinarily low amounts (Fig. 8A, top, leftmost lane). In addition, this polypeptide appeared to be of slightly lower molecular weight than the putative smaller, C-terminally truncated TSGA2 isoform, identified in the testis only by N (Fig. 6). Whether this novel isoform is an in vitro proteolytic artifact of the fractionation procedure or an in vivo posttranslationally processed product is presently unknown.
Because a portion of the 30-kDa isoform partitioned into the second cytoskeletal fraction by compartmental fractionation, and because at least some of this isoform appeared to remain in the sperm tail midpiece by immunofluorescence after removal of the mitochondrial sheath, we hypothesized that this portion of the 30-kDa isoform might be associated with the ODF, the axoneme, or both. Further, neither of our fractionation procedures offered any guarantee that the principal piece-specificity of the 40-kDa isoform post-Triton X-100 washes or mitochondrial sheath extraction resulted from localization of the overwhelming majority of this cytoskeletal polypeptide to the FS, especially because its detergent-solubility characteristics were clearly different from those of AKAP4 (Fig. 8A), and because its putative counterpart in C. intestinalis sperm tails appeared to be a component of the axonemal radial spoke complex. We therefore tested whether substantial portions of the 30- and/or 40-kDa mouse TSGA2 isoforms partitioned with either FS and/or ODF protein-enriched fractions by separating sperm tails from sperm heads, and extracting FS and ODF protein-enriched preparations from the tails [18, 27]. Western blot results from extracts enriched for FS or ODF proteins suggested that a majority of the 40-kDa isoform (65%–80% of the total isolated) was FS-associated (Fig. 8B) although a visible minority of this protein remained in the non-FS/non-ODF protein-enriched fractions (data not shown). Additionally, approximately equally abundant portions of the 30-kDa isoform partitioned with the FS and ODF protein-enriched fractions (Fig. 8B), whereas approximately 30%–40% of the solubilized total remained in the non-FS/non-ODF fraction (data not shown). Probing equivalent Western blots with anti-ODF2, anti-AKAP4, and anti-ßTubulin antibodies provided an estimate (generally <5%) of cross contamination of both the FS and the ODF protein-enriched fractions by components from other cytoskeletal structures of the sperm tail.
An 30-kDa TSGA2 Isoform is a De Novo Product of the Epididymis
Western blot analysis of mouse epididymal protein extracts with N and C demonstrated the presence of both the 40- and 30-kDa isoforms. As was the case for sperm, N identified only the larger isoform, whereas C recognized both; however, for both antibodies, the 40-kDa isoform was barely above background levels, and for C, the 40 kDa:30 kDa ratio was strikingly reduced relative to the same ratio in sperm protein lysates (Fig. 9A). One possible interpretation of these results was that a 30-kDa isoform of TSGA2 was expressed in large quantities de novo in the epididymis.
|
We derived additional evidence for this explanation from Western blot analysis of mouse epididymal protein extracts depleted of motile sperm by swim out. In that case, neither N nor C identified the 40-kDa polypeptide of apparent testicular origin, but C recognized an abundant 30-kDa polypeptide (Fig. 9B).
To determine whether an 30-kDa TSGA2 isoform was indeed a de novo product of the epididymis, we performed RT-PCR of epididymally-derived RNA with forward and reverse primers that flank the coding region of testicular Tsga2 mRNA (see Materials and Methods). On agarose gels putative Tsga2 cDNAs from the epididymides of tw5/tw32 and +/+ males appeared less abundant and slightly smaller than testicular Tsga2 cDNAs (Fig. 9C). Sequence analysis of the epididymally-derived cDNAs demonstrated that the second exon of testicular Tsga2 was precisely deleted in these novel transcripts (Fig. 9D). Interestingly, a review of orthologous sequences used in our initial alignment (Fig. 5) revealed that one (F; accession no. XP_514920) of the two (E and F) predicted chimpanzee TSGA2 isoforms carried a second exon deletion.
The coding regions of the epididymal Tsga2 cDNAs were each predicted to translate into polypeptides of 29.8 kDa. In addition, the inability of N to recognize these polypeptides was explained by the absence of 12 of 15 residues at the N-terminus of the deleted second exon against which this antibody was raised (Fig. 9D). Because N also did not detect the 30-kDa sperm isoform of TSGA2, these data suggested that the epididymal TSGA2 isoform and the 30-kDa sperm isoform were structurally similar, if not identical to each other. However, because the 30-kDa isoform found in sperm could be a posttranslationally processed product of the 40-kDa isoform, its origin is not presently known.
DISCUSSION
We set out to determine the basis of Ccub activity [6] via functional and genetic analyses of sperm from males heterozygous for three new heterospecific recombinant In(17)4 homologs of the t complex, SR12, SR14, and SR8, and the t haplotype, tw32. Initial results indicated a necessity for genetic redefinition of Ccub, and led us to propose a modified version of our original hypothesis concerning the molecular basis of curlicue [6]. This retooled hypothesis states that demonstration of the curlicue phenotype by >90% of the cauda epididymal sperm from a t/t male requires the homozygous expression of the t alleles of at least three, rather than two, genetic elements (Fig. 2B): a proximal factor, Ccua (presently believed to be Dnahc8 [7–10]), located proximal to the SR2 proximal M. spretus breakpoint and distal to the SR3 proximal M. spretus breakpoint; and a complex distal factor, Ccub, consisting of two synergic genes, Ccub1 and Ccub2, the former located distal to the SR2 proximal M. spretus breakpoint and proximal to the SR5 distal M. spretus breakpoint, and the latter, located between the SR14 proximal M. spretus breakpoint and the distal end of the t complex. Interestingly, the genomic interval to which Ccub1 maps is entirely coincident with the region circumscribing the proximal stop oolemma penetration factor, Stop1p [6]. Thus, it was possible that Ccub1 and Stop1p were identical. If so, a candidate Ccub1 protein would localize to the sperm tail and the sperm head.
We have also demonstrated that a model of Ccub/Stop1 properties and function should take into account breeding studies involving SR2 homozygous males (SR2 carries the + allele of Ccua, but the M. spretus alleles of Ccub and Stop1; see Fig. 2B and [6]) that show that although SR2/t males are sterile, males homozygous for SR2 are fertile. Thus, we have proposed that the M. spretus alleles of candidates for Ccub or Stop1 elements must be sufficiently wild-type in structure and expressed at sufficient levels when homozygous to promote male fertility, but when heterozygous with substantially mutated t alleles, they fail to complement either curlicue or stop, traits tightly linked to t/t male sterility, because they are haploinsufficiently expressed in the M. musculus genetic background.
Despite having demonstrated that the M. spretus and t alleles of Tsga2 meet the primary requirements of our model, we at first considered Tsga2 to be a weak candidate for either genetic factor, based on reports contending that TSGA2 was absent from mature mouse sperm [11, 12]. A further review of these earlier reports made it clear, however, that the assays employed to demonstrate the absence of TSGA2 from sperm were indirect at best, with neither study demonstrating said absence by probing Western blots of mouse sperm protein lysates with antisera raised against mouse TSGA2.
We employed two affinity-purified monospecific antisera raised against mouse TSGA2 N- and C-terminal peptide epitopes of suspected high antigenicity in Western blot studies to establish the presence of two major TSGA2 isoforms, an 40- and an 30-kDa polypeptide in mouse cauda epididymal sperm protein lysates (Figs. 6, 8, and 9). Additional studies of sperm by indirect immunofluorescence and protein fractionation/Western blotting suggested that both isoforms localize to the proximal fibrous sheath and are differentially distributed in the periacrosomal space along the anterior acrosome as well as the region between the plasma membrane and the outer membrane of the mitochondrial sheath. Additionally, a major portion of the smaller isoform, but not the larger one, appeared to localize to the outer dense fibers, and perhaps to mitochondria (Figs. 7 and 8).
Taken together, the results of our protein localization studies have enhanced the probability that TSGA2 isoforms may be components of the curlicue and/or stop pathways. However, the physiological mechanisms by which TSGA2 isoforms affect processes necessary for successful fertilization, and which may be perturbed by TSGA2t, remain unknown. The conservation of a tandem array of MORN repeats occupying the N-terminal part of all TSGA2 orthologous proteins implies that they are functionally significant components of TSGA2 isoforms. Homologous MORN repeats are integral components of junctophilins [28], several members of the phosphatidylinositol-4-phosphate 5-kinase family [29], and the histone-lysine N-methyltransferase, SET7/9 [30], as well as a number of other proteins. However, a definitive functional relationship between all of these proteins has yet to be established.
Recent findings from the Pfam database (http://www.sanger.ac.uk/cgi-bin/Pfam/interaction_domains.pl?acc=PF02493) indicate that MORN repeats are capable of interacting with each other and with other domains such as the SET domain at the C-terminus of SET7/9. In establishing these interactions, investigators determined domain-domain contacts by mapping Pfam domains onto PDB structures, followed by the identification of interdomain bonds. However, the algorithm employed did not distinguish between biological and crystal contacts. Thus, it is possible that at least some of these interactions are artifacts.
Nonetheless, it may be the case that the MORN repeats in TSGA2 isoforms modulate inter- and/or intramolecular interactions, thereby resulting in the formation of homo- and/or heteromers between MORN-containing molecules. However, the major mutations in the t alleles reside in the two acid-rich regions (ARRs) flanking the MORN repeats. The sequences of the ARRs are highly conserved in TSGA2 orthologs from all species studied to date (Fig. 5), a sign that they are likely to be involved in an activity of physiological significance, and would be affected adversely by the nonsynonymous nature of the mutations in the t alleles. Interestingly, the ARRs of TSGA2 show a striking homology to glutamic acid-rich regions in TITIN/CONNECTIN, recently shown to modulate calcium-induced conformational changes in adjacent tandem arrays of PEVK repeats, thereby controlling TITIN-based tension in muscle fibers [31]. Given the calcium-sensitive nature of the curlicue phenotype [3–10], the possibility that the TSGA2 ARRs are calcium-regulated modulators of MORN-MORN, MORN-SET, or other MORN interactions seems an attractive one.
Together with our TSGA2 localization results, the findings from previous studies concerning the location of TSGA2 orthologous proteins residing in the sperm tails of nonmammalian organisms may provide us with additional clues concerning the general function of TSGA2 isoforms in mammalian sperm. For example, the Tsga2 ortholog from Cyprinus carpio encodes a protein, MSAP, which localizes to the entire length of the sperm tail, including the centrosome [12]. Although mature mammalian sperm are devoid of an intact centrosome/basal body with which to anchor the proximal ends of the axonemal doublets, mammalian sperm anchor their doublets peripherally through the ODF, thus increasing transverse force production during initiation and propagation of flagellar bends [32]. Whether or not the mouse TSGA2 isoforms present in the ODF play an analogous role to the centrosomal MSAP isoform in carp sperm is yet to be determined, though it is an appealing prospect.
Additionally, the Tsga2 ortholog from C. intestinalis encodes a protein that localizes to the axonemal radial spoke complex, where it coimmunoprecipitates with radial spoke protein 3 (RSP3) [13], an AKAP thought to regulate the activity of inner arm dynein through protein phosphorylation and dephosphorylation [33]. Because the major protein of the mammalian sperm tail-specific FS is AKAP4 [34], it may be that mammalian TSGA2 isoforms localizing to the FS play analogous roles to their C. intestinalis counterparts to control principal piece-specific outer dynein arm activity [10, 35]. Given that the Ccua candidate, Dnahc8, codes for a unique, principal piece-specific outer dynein arm heavy chain [10], it is tempting to speculate that this is, indeed, the case. In light of these findings and our evidence that TSGA2 isoforms localize to many regions of the mouse sperm tail as well as to the sperm head, it seems likely that in mammalian sperm these proteins may serve as multifaceted adaptors in numerous calcium-regulated signaling pathways.
Finally, although the suggestion that the 30-kDa sperm isoform of TSGA2 is identical to the structurally similar isoform produced by the epididymis is enticing, Tsga2 shows no evidence of encoding a signal peptide that would allow the 30-kDa isoform to translocate from epididymal cells to sperm by typical processes during epididymal transit. Although it is possible that the epididymal isoform uses epididymosomal transfer to gain access to sperm [36], epididymosomes (membranous bleblike vesicles capable of delivering proteins without signal sequences from the epididymal principal cell epithelium to sperm) generally deliver their cargoes to defined domains on the sperm surface. In fact, only a single instance has been documented in which an epididymosomally transferred protein has been internalized by sperm [36, 37]. Interestingly, this single protein (macrophage migration inhibitory factor) is associated with the ODF of the sperm tail.
FOOTNOTES
1 Supported by NIH grants HD31164 and HD38359 (to S.H.P.). 
2 Correspondence: Stephen H. Pilder, Department of Anatomy and Cell Biology, Temple University School of Medicine, 3400 N. Broad St., Philadelphia, PA 19140. FAX: 215 707 2966; stephen.pilder{at}temple.edu
3 Current address: Department of Obstetrics and Gynecology, University of Pennsylvania Medical Center, Philadelphia, Pennsylvania 19104.
4 Current address: Department of Obstetrics and Gynecology, Chinese University of Hong Kong, Hong Kong, New Territory, People's Republic of China
Received: 24 July 2005.
First decision: 2 September 2005.
Accepted: 12 December 2005.
REFERENCES
-type axonemal dynein heavy chain isoform confined to the principal piece of the sperm tail. Dev Biol 2005 285:57-69[Medline]This article has been cited by other articles:
|
|
 |
|
  P. Puri, K. Myers, D. Kline, and S. Vijayaraghavan Proteomic Analysis of Bovine Sperm YWHA Binding Partners Identify Proteins Involved in Signaling and Metabolism Biol Reprod, December 1, 2008; 79(6): 1183 - 1191. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
