| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Biological Sciences, University of Warwick, Coventry CV4 7AL, United Kingdom
ABSTRACT
Proteins immunologically related to intermediate filaments have been identified in the sperm fibrous sheath but remain uncharacterized. We isolated and characterized a novel intermediate filament-related protein (FS39) localized to the fibrous sheath of the sperm tail. We used Northern blot analysis to establish that FS39 is transcribed predominantly in the testis of mice >18–20 days old. At this age, spermatogenesis has proceeded to the development of the first round haploid spermatids. In situ hybridization revealed that FS39 mRNA is first detectable in late step 3 spermatids, is at its highest level during steps 9 and 10, and diminishes in steps 13 and 14. Western blot analysis identified a single protein of 39 kDa in mouse and rat testis and epididymis, suggesting the protein is conserved in rodents. Indirect immunofluorescence localized FS39 to the fibrous sheath of the sperm tail, and in testis sections expression was detected from step 13 and step 14 spermatids onward, indicating FS39 is under translational control. Southern blot analysis showed FS39 to be a single copy gene, and hybridization to human genomic DNA suggested that a human equivalent gene is present. These results demonstrate that FS39 is transcribed in testis tissue during the haploid phase of spermatogenesis, is present in mature sperm, and codes for a novel 39-kDa intermediate filament-related protein of the fibrous sheath.
developmental biology, gene regulation, spermatid, spermatogenesis, testis
INTRODUCTION
The specialized organelles of the spermatozoon are formed during the haploid stages of spermatogenesis, also known as spermiogenesis [1]. Spermiogenesis in mice is divided roughly into processes occurring in round spermatids (steps 1–8) and those occurring in elongating spermatids (steps 9–16). These steps can be further subdivided into the Golgi phase (steps 1–3), the cap phase (steps 4–7), the acrosome phase (steps 8–12), and the maturation phase (steps 13–16) [2]. In round spermatids, the acrosome and flagellum start to differentiate [3, 4]. In elongating spermatids, the mitochondria become arranged in a spiral around the base of the flagellum, a fibrous sheath is deposited around the flagellar axoneme, and the nuclei become elongate, dense, and compact [4, 5]. Each of these processes involves the synthesis of testis-specific proteins, many of which function as structural components. These proteins have been characterized to various degrees, but many are known by only their molecular weights on the basis of subcellular structure analysis and the availability of antisera.
The accumulation of specific transcripts during the haploid phase of spermatogenesis has been demonstrated by differential screening and extensive Northern blot analysis [6]. Certain mRNAs are abundant in round spermatids, and these RNAs are transcribed exclusively in haploid cells [7]. An unusually high proportion of poly(A)+ RNA in the testis is nonpolysomal [8, 9], which implies a crucial role for translational gene regulation in male gamete development [10, 11]. The identification of mRNAs expressed in haploid cells is fundamental to understanding transcriptional, posttranscriptional, and translational regulation during spermiogenesis.
Mammalian sperm tails are characterized by their complex cytoskeletal structure [12]. The nine outer dense fibers that surround the microtubular axonemes are encompassed by a mitochondrial sheath in the midpiece and the fibrous sheath in the principal piece. The fibrous sheath itself is a unique sperm tail component and is a tapering cylinder that encases both the outer dense fibers and axoneme. It is composed of two cytoskeletal elements, longitudinal columns that are attached to microtubular doublets 3 and 8 and numerous regularly spaced transverse ribs that extend halfway around the sheath and connect the two columns [13]. Functionally, the fibrous sheath, which is characterized by its elastic rigidity, is thought to provide structural support to the sperm tail during its movement [14]. The longitudinal columns may determine the plane of bending by restricting the participation of doublets 3 and 8 in microtubule sliding and axoneme bending [15]. Infertile men with fibrous sheath defects have a major kinetic anomaly characterized by a decrease of the amplitude of lateral head displacement [16]. Infertility is attributed to impairment of sperm propulsion through the cervical mucus and zona pellucida [16]. The fibrous sheath also functions as a scaffold for enzymes that produce energy required for motility of sperm at the time of fertilization [17]. A-kinase anchoring protein (AKAP) 82 can tether cAMP-dependent protein kinase A and therefore sequester the kinase in close proximity to its physiological substrates [18]. The initiation and maintenance of sperm motility is regulated by a cascade of cAMP-dependent phosphorylation/dephosphorylation events [19]. The attachment of the longitudinal columns of the fibrous sheath to microtubular doublets of the axoneme may function to regulate the flagellar beat via this phosphorylation-dependent mechanism [15]. Research aimed at identification of components of the fibrous sheath has revealed the presence of several polypeptides. For example, the rat fibrous sheath consists of at least 20 polypeptides [20], some of which appear to have homologues in human and mouse, suggesting that they are conserved among mammalian species [21–23]. Fibrous sheath proteins characterized to date include AKAP82 [18, 24], testis-specific AKAP80 (TAKAP80) [25], three enzymes (glutathione-S-transferase [26], glyceraldehyde-3-phosphate dehydrogenase [27], and type I hexokinase [28]), fibrous sheath 95 (FS95) [29], FS75 [30], ropporin [31], and rhophilin [32].
In our previous studies, the differential display technique was utilized to identify testicular mRNAs expressed during the haploid stages of mouse spermatogenesis [33, 34]. This approach resulted in the identification of a differentially expressed cDNA designated FS39, previously named differential display clone 8 (accession number Y09878). Database searches using the nucleotide and amino acid sequence suggested that FS39 coded for a novel protein with similarities to filamentous proteins such as myosin [34]. Here, we present the expression pattern and characterization of FS39 as a novel intermediate filament-related protein of the sperm fibrous sheath.
MATERIALS AND METHODS
Isolation of Total RNA
CD1 mice were obtained from Charles River Laboratories (Margate, UK). Human testes obtained from patients with prostate cancer (Walsgrave Hospital, Coventry, UK) were morphologically normal when examined under light microscopy. Total RNA was isolated using the SV total RNA isolation system according to the manufacturer's instructions (Promega, Southampton, UK).
Purification of Poly(A)+ RNA
Poly(A)+ RNA was purified from total RNA using biotinylated oligo(dT) and streptavidin paramagnetic particles with the poly(A)Tract mRNA isolation system (Promega).
Northern Analysis
Poly(A)+ RNA samples were electrophoresed as previously described [35] on 1.2% gels, except samples were denatured by heating to 65°C for 10 min in 20 mM guanidine thiocyanate and transferred onto nylon filters (Hybond-N, Amersham, Bucks, UK) for 16 h [36]. RNA blots were hybridized as previously described [37] to full-length [
32P]dCTP random-primed FS39 cDNA.
Genomic DNA Isolation
DNA was isolated from mouse kidney tissue and human testis tissue as previously described [37].
Southern Analysis
Southern blots were performed using 20 µg of digested genomic DNA as previously described [37]. The mouse genomic DNA blot was hybridized to full-length FS39 [32P]cDNA. The human blot was hybridized with a 1.4-kilobase (kb) [32P]cDNA representing nucleotides 423–1807 of FS39. The blots were washed once in 2x standard saline citrate (SSC), 0.1% SDS for 20 min at room temperature and twice in 0.2x SSC for 20 min at 65°C (mouse blot) or 0.2x SSC for 20 min at room temperature (human blot).
In Situ Hybridization
Sense and antisense RNA probes were synthesized by in vitro transcription of pBluescript vectors harboring the clone 85 insert with T3 or T7 RNA polymerase after linearization with XhoI or EcoRI, respectively. Clone 85 represents approximately 550 base pairs (bp) of the 3' end of the FS39 cDNA [34]. Probes were labeled by incorporation of digoxigenin (DIG) uridine triphosphate (DIG-UTP; Boehringer, Lewes, UK) and hydrolyzed to 
300 base fragments.
Hybridization was carried out by a modified method [38] to histological sections of testes from sexually mature CD1 mice. Testes were fixed in 4% paraformaldehyde, dehydrated through an ethanol series, embedded in paraffin, sectioned at 6 µm, mounted on glass slides, and hydrated. Sections were digested with 1 µg/ml proteinase K (Sigma, Gillingham, UK) in 0.1 M Tris HCl, pH 8.0, 50 mM EDTA for 60 min at 37°C, treated with 0.25% (v/v) acetic anhydride, and dehydrated to absolute ethanol. Sections were prehybridized in hybridization buffer minus dextran sulfate for 60 min at 45°C. The probe (20 ng/section) was heated to 80°C for 3 min in hybridization buffer: 50% formamide, 10 mM Tris HCl, pH 8.0, 1 mM EDTA, 1x Denhardt reagent, 500 µg/ml Escherichia coli tRNA (type XX; Sigma), and 10% (w/v) dextran sulfate. The mixture was chilled and applied to each section (20 µl), and sections were covered with a strip of Parafilm. Slides were incubated in a humid chamber at 45°C for 16 h, washed in 2x SSC for 30 min at room temperature, treated with 100 µg/ml pancreatic RNase A (Sigma) for 60 min at 37°C, and washed in 0.5x SSC for 30 min at 50°C.
Detection of DIG-labeled RNA probes was performed using an anti-DIG-alkaline phosphatase conjugate according to the manufacturer's instructions (Boehringer). To identify cell types, sections were stained with Harris hematoxylin (Sigma) for 30 min, rinsed in water, and then stained for 30–60 sec with eosin Y (BDH, Poole, UK). Sections were then mounted with D.P.X. (BDH) and photographed with Kodak Ektachrome 160T film (Kodak, Rochester, NY) using a neutral blue filter.
In Vitro Transcription
In vitro transcription was carried out with 10 µg of linearized plasmid DNA in 100 µl containing 1x transcription buffer (Promega), 2.5 mM rNTPs (Promega), 80 U ribonuclease inhibitor (Gibco BRL, Paisley, UK), 0.5 U yeast inorganic pyrophosphatase (Sigma), and 200 U T3 or T7 RNA polymerase (Gibco BRL) and incubated for 2 h at 37°C. RNA was then extracted twice with phenol:chloroform (1:1) and once with chloroform and then precipitated with ethanol.
In Vitro Translation
In vitro translations were performed with 100 ng of RNA in a 25-µl reaction containing 17.5 µl of rabbit reticulocyte lysate (Promega), 1 µl amino acid mixture (minus methione; Promega), 20 U RNasin ribonuclease inhibitor (Gibco BRL), and 20 µCi [35S]methionine (Amersham) and incubated at 30°C for 60 min. Products were then analyzed on a 10% SDS-polyacrylamide gel and exposed to x-ray film (Fuji, Tokyo, Japan).
Cell Culture
The insect cell lines (Spodoptera frugiperda) Sf9 and Sf21 were propagated at 27°C in Grace medium (Gibco BRL) supplemented with lactalbumin hydrolysate (3.3 g/L) and yeastolate (3.3 g/L) (Difco, Detroit, MI) and containing 10% fetal calf serum (FCS). The High Five cell line (Invitrogen, Leek, The Netherlands) was propagated at 27°C in serum-free Ex-Cell 401 medium (JRH Biosciences, Cambridge, UK). During infection with the baculovirus, the cells were cultured in the above medium with the addition of penicillin and streptomycin (10 µg/ml).
Construction of Recombinant Baculovirus Vector
To construct the recombinant baculovirus vector, the coding region of FS39, which includes nucleotides 886–1807, was removed from pBluescript-FS39 by digestion with EcoRI and BglII. This fragment was subcloned into pUC28 EcoRI/BamHI sites and subsequently digested with EcoRI and SalI for subcloning into the EcoRI/SalI sites of pBlueBacHis2B baculovirus transfer vector (Invitrogen). Constructs were sequenced using Sequenase 2.0 (Amersham) to ensure that the cDNA had been ligated correctly in frame with the translational start site.
Construction and Purification of Recombinant Virus
Recombinant transfer vector and Bac-N-Blue viral DNA (Invitrogen) were cotransfected into Sf21 cells using Insectin Plus liposomes (Invitrogen). Recombinant virus was identified using a plaque assay by infecting cell monolayers with dilutions of the transfection stock and isolating recombinant plaques from an agarose overlay using color selection with X-gal. Recombinant plaques (85%) were examined for wild-type virus contamination by polymerase chain reaction analysis of the viral DNA. Pure recombinant virus was then used in the preparation of high-titer viral stocks.
Detection and Expression of Recombinant Protein
Purified recombinant virus was used to infect monolayers of cells with a multiplicity of infection of 10. At 0, 24, 36, 48, 60, 72, 84, and 96 h postinfection, cells were detached by mechanical disruption. Cells were pelleted, resuspended in 20 mM sodium phosphate buffer, pH 7.0, and lysed by four freeze/thaw cycles. Protein concentrations were determined using colorimetric protein assay (BioRad, Hemel Hempstead, UK), and equal amounts of protein were then electrophoresed on a 10% SDS-polyacrylamide gel for analysis by Western blotting.
Immunoblotting
Proteins were electroblotted on nitrocellulose (Hybond C; Amersham), and nonspecific binding was blocked using 3% BSA. The blot was incubated for 2 h with primary antibody and washed twice in Tris-buffered saline with Tween-20: 20 mM Tris HCl, pH 7.5, 500 mM NaCl, and 0.05% Tween-20. Bound antibody was detected using an alkaline phosphatase-conjugated secondary antibody. The blot was washed as above and equilibrated in 100 mM Tris HCl, pH 9.5, 100 mM NaCl, and 50 mM MgCl2, and 75 mg/ml nitroblue tetrazolium chloride (NBT) and 50 mg/ml 5-bromo-4-chloro-3-indolyl-phosphate (BCIP) were used for chromogenic detection.
Tunicamycin Treatment
Sf21 cells were infected with FS39 recombinant baculovirus and incubated for 2 h at 28°C. Culture medium was then replaced by medium containing 1 µg/ml of tunicamycin (Sigma). Infected cells were incubated for 3 days at 27°C, and cell lysates were analyzed by Western blot assays.
Purification of Recombinant FS39 Protein
Purification under native conditions showed FS39 to be highly insoluble; therefore, cell pellets from the large-scale production of FS39 recombinant protein were purified using Probond resin (Invitrogen) according to the manufacturer's instructions. Binding and elution of recombinant FS39 were monitored by analyzing fractions during purification by immunoblotting with the anti-Xpress antibody. The yield of recombinant protein purified from 1 x 109 High Five cells was approximately 6 mg.
Antibody Production
For antibody preparation, a rabbit was injected s.c. at four sites with 500 µg of purified protein emulsified with an equal volume of Freund complete adjuvant. After 4 wk, a booster of 500 µg of purified protein emulsified with an equal volume of Freund incomplete adjuvant was administered. Blood from the rabbit was collected from the marginal ear vein, and the antiserum (anti-FS39) was characterized by immunoblotting.
Indirect Immunofluorescence
Mouse sperm prepared from the epididymis by the swim-up method were smeared on a microscope slide, fixed in a mixture of methanol and acetone (1:1), and air dried. Mouse testis was fixed in 4% paraformaldehyde and sectioned in paraffin at 7 µm. Slides were treated with 1% Triton X-100 for 60 min and washed twice in 1% BSA in PBS. Nonspecific binding was blocked by incubation for 20 min using 10% FCS in PBS and washed as above. Slides were incubated for 1 h with anti-FS39 serum (1:100), washed, incubated for 45 min in the dark with an anti-rabbit fluorescein isothiocyanate (FITC) antibody (Sigma), washed, stained with 4',6'-diamidino-2-phenylindole (DAPI), and mounted with Vectorsheild mounting medium (Vector, Burlingame, CA). Fluorescence was viewed with a Nikon (Tokyo, Japan) photomicroscope equipped with epifluoresence, FITC, and DAPI filters, and sections were photographed with Kodak T-Max P1600 film.
Protein Sequence Analysis
Protein database searches were performed using BLAST [39], Fasta [40], and PROPSEARCH [41]. BLAST and Fasta use linear alignment methods, whereas PROPSEARCH uses a nonlinear comparison based on amino acid composition. Secondary structure was predicted using the consensus secondary structure program [42] that compares results from 12 secondary structure prediction programs. Coiled-coil regions were predicted with the coiled-coils program [43]. The amino acid sequence was examined for motifs by comparisons with the PROSITE database [44].
RESULTS
Northern Analysis
Tissue specificity of FS39 was determined by Northern blot analysis using 5.0 µg of poly(A)+ mRNA from the following mouse tissues: testis, ovary, lung, liver, brain, epididymis, kidney, skeletal muscle, and spleen. Three transcripts of 1.2 kb, 2.4 kb, and 3.6 kb were identified only in testis tissue after 24 h of exposure; three other transcripts in liver tissue could also be faintly seen. After 4 days of exposure, the three liver transcripts of approximately 1.5 kb, 2.5 kb, and 5.0 kb could be identified more clearly. None of these bands were present in any other tissue (Fig. 1A). Developmental expression of FS39 was determined by Northern blot analysis using 2.6 µg of poly(A)+ RNA from mouse testes 15–16, 18–20, 24–26, and 60 days old and from human testis. The autoradiograph (Fig. 1B) displayed a band of 2.4 kb appearing in mouse testis 18–20 days old, with the signal increasing in intensity to maturity at 60 days of age. After exposure to x-ray film for 14 days, a second band of approximately 3.6 kb and a faint third band of 1.2 kb (within the smear from the 2.4-kb band, better seen in Fig. 1A) was identified from testis of mice 18–20 days of age and older. No bands were present in mouse testis 15–16 days of age or in the human testis.
|
Southern Analysis
Southern blots were used to determine the number of copies of the FS39 gene in the mouse genome and to detect a human homologue. DNA digested with the restriction enzymes HindIII and BamHI yielded only one high molecular weight band (HindIII: 15 kb and 3.5 kb; BamHI: 12 kb). EcoRI-digested DNA yielded two bands (14 kb and 9.4 kb); FS39 cDNA contains an EcoRI recognition site at 885 bp. Hybridization of the human blot with the 1.4-kb coding region of FS39 under reduced wash stringencies produced several bands (HindIII: 9 kb and 4.7 kb; EcoRI: 7.2 kb and 5.6 kb; BamHI: 7.5 kb and 6 kb), suggesting that a human homologue of FS39 may exist (Fig. 2).
|
In Situ Hybridization of Mouse Testis Sections
In situ hybridization was used to identify cells containing FS39 transcripts in paraffin-embedded sections of mature mouse testis. Using an antisense FS39 probe, signal was detected only in the haploid cells of the testis (Fig. 3A). Expression was first detected in round spermatids from steps 3 and 4, peaked around steps 9 and 10, and thereafter declined such that by steps 15 and 16 of maturing spermatids no signal was detected (Fig. 3, B–E). The FS39 sense probe acted as a negative control (Fig. 3F). To identify cell types after in situ hybridization, testis sections were stained with hematoxylin and eosin, which highlighted the signal. The eosin preferentially stained areas where the precipitate had formed from the reaction of the alkaline phosphatase with the BCIP/NBT color substrate. This staining was increased above background in proportion to the intensity of the signal produced after in situ hybridization (Fig. 4) and conveniently permitted the visualization of both signal and cell types simultaneously.
|
|
In Vitro Translation of FS39 RNA
The largest open reading frame (nucleotides 288–1887) codes for a 533-amino acid protein with a molecular mass of 62 kDa [34]. To verify the translational start site, a pBluescript SK vector containing FS39 was used for the synthesis of RNA by in vitro transcription. The RNA was then translated in a cell-free system and analyzed by SDS-PAGE. The protein product was 36 kDa, indicating that translation begins at the fourth in-frame start site. The 36-kDa product is produced from the open reading frame at nucleotides 867–1887, which codes for a 340-amino acid protein with a calculated molecular mass of 39.5 kDa (Fig. 5).
|
Baculovirus Protein Expression
The baculovirus expression vector was constructed as described to express the 39.5-kDa protein product from nucleotides 867–1887. Pure recombinant virus for both FS39 and CAT (control) were used in initial time course experiments to determine the level of expression. Sf9 and High Five cell lines showed approximately twice the level of expression of that of the Sf21 cell line (Fig. 6A). In all cell lines, FS39 recombinant was expressed as a doublet of approximately 44 kDa and 48 kDa, whereas Sf9 and Sf21 cells produced an equal ratio of these proteins, and High Five cells predominately expressed the 48-kDa protein. Both recombinant proteins are larger than the 40.6-kDa predicted product, which suggested that FS39 undergoes some sort of posttranslational modification. To test whether glycosylation was involved, tunicamycin was added to the medium after infection. Samples examined by Western blot still contained the doublet, suggesting that the modification is not N-glycosylation (Fig. 6B).
|
Western Blot Analysis
Western blot analysis was performed on extracts prepared from human, rat, and mouse testis, rat and mouse epididymis, and human sperm (Fig. 7). Antiserum raised against FS39 recombinant recognized a protein of approximately 39 kDa in both mouse and rat testis and epididymis. The human samples failed to show any specific bands.
|
Indirect Immunofluorescence
Using antiserum against FS39 recombinant protein, sperm prepared from mouse epididymis was analyzed. Immunoreactivity was seen only in the principal piece of the sperm tails. Fluorescence was greatest at the proximal end of the principal piece and became more diffuse towards the distal end (Fig. 8). Under higher magnification, the staining was evident as two parallel lines of striated structures running along the principal piece. Treatment of sperm with Triton X-100, which solubilises the plasma membrane [23], increased the intensity of fluorescence, whereas treatment with dispase (4 U/ml for 10 min at 37°C; data not shown) completely abolished the signal. Tissue sections from mature mouse testis showed expression in the tubule lumen from step 13 and step 14 spermatids as the fibrous sheath is constructed. Fluorescence increases in step 15 and step 16 spermatids as the fibrous sheath framework is completed (Fig. 9).
|
|
DISCUSSION
Mammalian spermatozoa are composed of specialized cytoskeletal elements, which appear to have no structural or protein counterparts in somatic cells. Most evident are the outer dense fibers, fibrous sheath, mitochondrial capsule, perinuclear theca, and acrosomal matrix. FS39 appears to have no apparent homologues, although extensive use of databases and protein prediction programs can suggest a possible role. FS39 possesses characteristics of fibrous sheath proteins such as high insolubility, is rich in aspartic acid and glutamic acid [45], and contains putative phosphorylation sites that may play a key role in both assembly and function of the fibrous sheath [46, 47]. Secondary structure prediction analysis revealed a rod domain of 276 amino acids with two helical regions that contained heptad repeats (a, b, c, d, e, f, g), in which residues a and d are generally hydrophobic. These regions are predicted to form coiled coils capable of dimerization. The N-terminal extension is hydrophilic, with a high probability of being exposed on the surface of the protein, and could therefore protrude from the filament backbone and confer its function. The general structure of the FS39 is very similar to the substructure of intermediate filaments (Fig. 10). Intermediate filament proteins have a common structural core of 300–330 amino acids that are flanked by nonhelical extra amino and carboxy terminal domains. The common core is composed of a heptad repeating unit where hydrophobic residues create a coiled-coil dimer, and the formation of such a structure is central to the function of all intermediate filament proteins [48]. The uniformity of this rod domain and two conserved motifs at the beginning (LNDR) and end (TYRKLLEGE) of this domain are critical to intermediate filament assembly [49]. FS39 has a rod domain 24 amino acids shorter than the consensus rod domain but maintains the same structural plan except for helix 1B, which departs from the consensus structure. The conserved motifs are not maintained at the beginning of the rod but show 44% identity at the end, which is conserved among species (Fig. 11). The termini are not homologous and confer a specific function on the molecule [48]. The termini can be of various lengths, although FS39 lacks a C-terminal extension. This lack is rare but not unique; other intermediate filaments such as phakinin a lens-specific protein also lack a C-terminal domain [50]. The fibrous sheath is composed of 5- to 6-nm-wide filaments [51], which could explain why FS39 has diverged from the consensus intermediate filament model that produces 10-nm filaments. Other intermediate filament-related proteins, such as filensin and phakinin, which form a beaded filament, have also diverged from the normal consensus rod domain to accompany different structural features [52]. Comparison of the 276-amino acid rod domain to the Swissprot database revealed low identities to filamentous proteins, all belonging to the class of fibrous proteins, such as nonmuscle myosin (Table 1). When multiple alignments were constructed, there was no consensus sequence evident although similarities were between helical domains. The FS39 amino acid sequence was used to search the expressed sequence tag (EST) database using the tblastn program. This program compares a protein query sequence against a nucleotide sequence database translated in all reading frames. The comparison revealed a rat EST (BE101971) and three human ESTs (AW899572, AI867422, AI651743), which when aligned formed a continuous sequence. When the nucleic acid sequences are aligned, they reveal 91% homology between mouse and rat sequences and 65% homology between mouse and human. All sequences contain poly(A) tails with identical length 3' untranslated region (UTR), stop codon, and polyadenylation signal (Fig. 11A). The putative glycosylation sites in the mouse are not conserved in rat and human, which provides further evidence that glycosylation may not occur, because it is not required functionally between species. Translated EST sequences revealed homologies of 88% and 54% between mouse and rat and between mouse and human, respectively (Fig. 11B). The highest homology is over the coiled coil domains of the helical structures, although only partial sequences could be compared. This homology is more evident when comparing the predicted secondary structure between proteins where these domains are seen to be highly conserved (Fig. 11C). The head domain sequence showed no identity to any Swissprot entry.
|
|
|
FS39 is expressed predominantly as a 2.4-kb transcript, implying that approximately 400 nucleotides of 5' UTR are to be found upstream of the known FS39 sequence. We assume that the 400-nucleotide section does not contain coding sequence; stop codons in all three reading frames have been found in the known 5' UTR. FS39 is expressed at high levels in the testis but at very low levels in the liver. In both tissues, the FS39 probe detected three transcipts; at present it is unknown whether FS39 is differentially spliced or whether these transcripts code for identical or novel proteins. Expression of FS39 mRNA commences 18–20 days after birth, which correlates with the development of the postmeiotic round spermatids. We previously showed by ribonuclease protection assay analysis that no expression was detected in either mouse Sertoli or Leydig cell lines [34], which suggests that expression is confined to the haploid germ cells of the testes. In situ hybridization confirmed that FS39 mRNA is germ cell specific and is expressed in spermatid steps 3–14, peaking during the acrosome phase (steps 8–12) of spermiogenesis.
The FS39 cDNA probe failed to hybridize to human testis RNA at high hybridization and wash stringencies. Under the conditions used, a similarity of more than 95% would be needed for the probe to hybridize. At reduced wash stringencies, a signal was detected in human genomic DNA. Some of the observed bands may represent nonspecific binding under low-stringency conditions, although primary sequencing of cDNA clones isolated from a human testis cDNA library under the same conditions show 95–97% identity to the human EST homologs (unpublished data). Southern blots indicate that FS39 is a single-copy gene and that a similar gene is present in humans. Although immunoblotting failed to detect FS39 in the human testis, the human protein sequence has diverged by 46% from the central linker to the C-terminus, which could explain why the antibody failed to cross-react. Failure of cross-species reactivity of fibrous sheath antibodies has previously been observed [53, 54].
The expression pattern of FS39 mRNA is similar to that of other haploid expressed mRNAs that are stored as free mRNPs for several days and translated in elongating spermatids after the cessation of transcription [9] due to nuclear condensation in step 9 spermatids [55, 56]. FS39 mRNA is subject to translational regulation; the protein is first detected at steps 13 and 14 during the maturation phase of spermiogenesis. Other proteins of the fibrous sheath, such as FS75, are subject to translational regulation [30]. Although fluorescence is first seen in steps 13 and 14, it is only faint in the sperm tails. The primary flagellum is so thin that ribosomes and organelles are excluded from it, implying that proteins destined for the maturing flagellum must be synthesized in the spermatid cell body and then transported to their sites of assembly [57]. Thus, FS39 may be translated in earlier steps. Although no fluorescence was detected in the cell bodies, mRNA levels do decrease in step 11 spermatids, which may be as a consequence of translation. AKAP82 is synthesized as an inactive precursor in the cell body that is transported down the axoneme to its site of assembly, where it is proteolitically cleaved into the mature protein [58]. FS39 contains a predicted amidation site at amino acid 11. Amidation sites are found at cleavage sites in precursor proteins, and FS39 could therefore undergo a similar processing to transport it to its site in the fibrous sheath. Immunoblot analysis revealed a protein of 39 kDa in mouse and rat testis and epididymis, suggesting that the protein is highly conserved in rodents. A protein of 37.2 kDa abundant in aspartic and glutamic acid has previously been identified by SDS-PAGE in the rat fibrous sheath [20]. The result of the immunoblot indicated that the translational start site in vivo might be identical to that in vitro; the products are very similar in size, 39 kDa and 36 kDa, respectively. The discrepancy in size is most likely due to posttranslational modification, as indicated by the two bands in the baculovirus expression system, but that does not occur in the in vitro translation system. FS39 is therefore translated from the 11th downstream ATG codon and possesses a long and problematical 5' UTR. This situation produces two questions: why do ribosomes bypass three in-frame start sites, two of which have a strong translational consensus sequence [59], and how does translation initiate at the fourth downstream start codon? A growing number of eukaryotic mRNAs that contain problematical 5' UTRs and are translated by a scanning independent mechanism have been reported [60]. The 5' UTR of FS39 is currently being investigated to see whether the current scanning mechanism for translation [61] applies to FS39 mRNA.
FS39 is distributed in the longitudinal columns of the fibrous sheath in a pattern identical to that of AKAP82 [18]. AKAP82 anchors protein kinase A to the fibrous sheath, and in response to cAMP the catalytic subunit of the kinase is released and becomes free to phosphorylate its substrate [47]. Initiation and maintenance of sperm motility is dependent on phosphorylation of flagellar proteins; however, the target proteins within the fibrous sheath have not been identified [19]. The predicted structure of FS39 together with posttranslational modifications and putative phosphorylation sites may well make FS39 a target of these phosphorylation events. Proteins immunologically related to intermediate filaments have been identified in the fibrous sheath [22, 23, 54]; however, this is the first reported case of the identification and characterization of such a protein. It has been suggested that spermatids contain their own unique intermediate filaments and associated proteins [23], and FS39 may well be one of these proteins, taking into account its localization and homology to filamentous proteins. Current studies involving protein-protein interaction will help elucidate the role of the fibrous sheath in normal sperm motility and of flagellar abnormalities that lead to male infertility, in particular in men with dysplasia of the fibrous sheath, whose sperm have severe structural abnormalities and are immotile [62]. This pathology appears to be familial, making it likely that a genetic component is involved [63]. Whether mutations in FS39 are involved is currently being investigated.
ACKNOWLEDGMENTS
We thank Debbie Clements (University of Warwick, Coventry, UK) for assistance with the baculovirus expression system, Rob Blacklock (Walsgrave Hospital, Coventry, UK) for the human testicular samples, and Richard Kennedy (Walsgrave Hospital) for organizing financial support from the GOAL Charity Fund.
FOOTNOTES
First decision: 11 December 2000.
1 This study was supported by the GOAL Charity Fund from the Assisted Conception Unit, Walsgrave Hospital, Coventry, UK, an NHS (West Midlands, UK) R&D research grant, and the Diabetes Medical Research Fund, Metabolic Unit, Coventry and Warwickshire Hospital, Coventry, UK. 
2 Correspondence. FAX: 44 2476 572748; rcatalano{at}cell.bio.warwick.ac.uk
Accepted: March 5, 2001.
Received: November 2, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  H.-J. Shi, A. Z. Wu, M. Santos, Z.-M. Feng, L. Huang, Y.-M. Chen, K. Zhu, and C.-L. C. Chen Cloning and Characterization of Rat Spermatid Protein SSP411: A Thioredoxin-Like Protein J Androl, July 1, 2004; 25(4): 479 - 493. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. H. Modarressi, B. Behnam, M. Cheng, K. E. Taylor, J. Wolfe, and F. A. van der Hoorn Tsga10 Encodes a 65-Kilodalton Protein That Is Processed to the 27-Kilodalton Fibrous Sheath Protein Biol Reprod, March 1, 2004; 70(3): 608 - 615. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Miranda-Vizuete, K. Tsang, Y. Yu, A. Jimenez, M. Pelto-Huikko, C. J. Flickinger, P. Sutovsky, and R. Oko Cloning and Developmental Analysis of Murid Spermatid-specific Thioredoxin-2 (SPTRX-2), a Novel Sperm Fibrous Sheath Protein and Autoantigen J. Biol. Chem., November 7, 2003; 278(45): 44874 - 44885. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. R. Brown, K. Miki, D. B. Harper, and E. M. Eddy A-Kinase Anchoring Protein 4 Binding Proteins in the Fibrous Sheath of the Sperm Flagellum Biol Reprod, June 1, 2003; 68(6): 2241 - 2248. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Yang, V. Chennathukuzhi, K. Miki, D. A. O'Brien, and N. B. Hecht Mouse Testis Brain RNA-Binding Protein/Translin Selectively Binds to the Messenger RNA of the Fibrous Sheath Protein Glyceraldehyde 3-Phosphate Dehydrogenase-S and Suppresses Its Translation In Vitro Biol Reprod, March 1, 2003; 68(3): 853 - 859. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Maier, S. Medrano, S. B. Sleight, P. E. Visconti, and H. Scrable Developmental Association of the Synaptic Activity-Regulated Protein Arc with the Mouse Acrosomal Organelle and the Sperm Tail Biol Reprod, January 1, 2003; 68(1): 67 - 76. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Yu, R. Oko, and A. Miranda-Vizuete Developmental Expression of Spermatid-Specific Thioredoxin-1 Protein: Transient Association to the Longitudinal Columns of the Fibrous Sheath During Sperm Tail Formation Biol Reprod, November 1, 2002; 67(5): 1546 - 1554. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Jimenez, R. Oko, J.-A. Gustafsson, G. Spyrou, M. Pelto-Huikko, and A. Miranda-Vizuete Cloning, expression and characterization of mouse spermatid specific thioredoxin-1 gene and protein Mol. Hum. Reprod., August 1, 2002; 8(8): 710 - 718. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Miyagawa, H. Tanaka, N. Iguchi, K. Kitamura, Y. Nakamura, T. Takahashi, K. Matsumiya, A. Okuyama, and Y. Nishimune Molecular cloning and characterization of the human orthologue of male germ cell-specific actin capping protein {alpha}3 (cp{alpha}3) Mol. Hum. Reprod., June 1, 2002; 8(6): 531 - 539. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||
