| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
research-article |
Section of Experimental Endocrinology,4 Department of Pharmacology, Universidade Federal de São Paulo-Escola Paulista de Medicina, São Paulo 04044-020, Brazil
Departments of Pediatrics5 and Cell Biology and Anatomy,6 University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-7500
ABSTRACT
Beta-defensins are small cationic peptides exhibiting broad spectrum antimicrobial properties. In humans, many beta-defensin genes are located within a cluster on chromosome 8p23. The sperm associated antigen 11 (SPAG11) gene is contained in this cluster and is unusual among the human beta-defensins due to its complex genomic structure and mRNA splicing pattern. Here we report the genomic organization of the Bos taurus SPAG11 gene located on chromosome 27q1.2, within a cluster of beta-defensin genes. The exon structures of the fused bovine SPAG11 gene and of the mosaic transcripts initiated at both A and B promoters were established, including identification of novel exons and transcripts not previously found in primate or rodent. Evolutionary analysis against primate, rodent, canine, and porcine orthologs was performed. In adult bulls SPAG11C, SPAG11E, and SPAG11U mRNAs were detected predominantly in the male reproductive tract, while SPAG11D transcript was detected in reproductive and nonreproductive tissues and SPAG11V and SPAG11W mRNAs were confined to testis. Differential expression of all six transcripts was observed in tissues from fetal and adult bulls, suggesting that similar mRNA splicing mechanisms govern SPAG11 gene expression during pre- and postnatal development. Immunolocalization of SPAG11C and SPAG11D/E was demonstrated in the epithelium of the epididymis and testis, and SPAG11D in association with epididymal spermatozoa. Recombinant full-length SPAG11D protein was strongly antibacterial, while the SPAG11E C-terminal peptide that contains the beta-defensin motif in its structure was somewhat less potent. Taken together, the results suggest that SPAG11 isoforms perform both immune and reproductive functions in cattle.
antibacterial, bovine, defensin, epididymis, gene expression, gene regulation, male reproductive tract, SPAG11
Defensins are antimicrobial molecules that contribute to innate host defense against bacteria, fungi, and viruses. They are small cationic peptides (18–45 residues) containing a well- conserved six-cysteine motif stabilized by three intramolecular disulfide bonds. Three mammalian defensin subfamilies (
-, ß-, and 
-defensins) have been classified, differing in their disulfide bond pairing, secondary structure, genomic organization, and tissue distribution. Both - and ß-defensin peptides occur in monkeys, humans, rodents, and other mammals [1–3]. The -defensins are expressed in mammalian polymorphonuclear leukocytes, mononuclear cells (including NK and T cells), small intestine, and genitourinary tract [4, 5]. ß-defensins, the oldest defensin subfamily, have been found primarily in epithelial tissues in contact with the environment, such as the airways, skin, intestine, and reproductive tract of mammals and invertebrates [1, 6]. Expression of a broad range of ß-defensins (and some -defensins) by various cells in the male reproductive tract of different species and on sperm surfaces in the epididymal lumen and in ejaculate have also suggested that these defensins have a role beyond innate immunity participating in events related to male fertility, sperm survival in the male reproductive tract, and sperm-egg interaction [7–11].
In humans, many ß-defensin genes are located within a cluster on chromosome 8p23 [6, 12, 13]. The SPAG11 (sperm associated antigen 11) gene, also known as EP2 (epididymal protein 2) in the monkey, HE2 (human epididymis 2) in the human, and Bin-1b (Spag11e) in the rat, is contained in this cluster and is unusual among the human ß-defensins because of its complex genomic structure and mRNA splicing pattern [10, 14]. Considerable progress also has been made in identifying complete SPAG11 genes in several other species, including chimpanzees, rhesus monkeys, mice, and rats [10, 15–17]. Different from the classical defensin genes in human and other primates, SPAG11 is considered to be a single gene derived from two ancestrally independent ß-defensin genes joined by read-through transcription. Primate SPAG11 gene expression is governed by this read-through transcription together with promoter choice (promoters A and B) and species-specific exon recruitment mechanisms that result in the production of at least 20 different alternatively spliced mRNAs [10, 15]. Transcripts include those derived from exons downstream of both promoters (i.e., SPAG11A and SPAG11D), as well as independent transcripts containing exons adjacent to only one promoter (i.e., SPAG11C and SPAG11E) [8, 14, 18–21].
The expression of these combinatorial transcripts is most abundant in the male reproductive tract and is determined by species as well as by organ-specific factors [10]. In rodents, expression of Spag11e and Spag11c mRNAs were reported in mouse epididymis [13], while Spag11c, Spag11e (also called Bin1b), and Spag11t transcripts were characterized in rat epididymis [16, 17, 22]. Orthologs of the other 17 SPAG11 transcripts found in primates have not been described in rodents. Each of the human and rhesus monkey (SPAG11A, SPAG11B, SPAG11C, SPAG11D, SPAG11E, SPAG11G, SPAG11K, SPAG11L) and rat (SPAG11C, SPAG11E, and SPAG11T) SPAG11 protein isoforms tested thus far has demonstrated potent antibacterial activity against Escherichia coli [10, 17, 23, 24]. In addition, major disease-causing organisms, such as Neisseria gonorrhoeae and Staphylococcus aureus, are greatly reduced by the C-terminal peptide of SPAG11A [25].
In a combination of activities unique to male tract host defense proteins, SPAG11 isoforms and other ß-defensins not only kill bacteria but also interact with spermatozoa [8, 11, 19], affecting motility [26] and zona-pellucida recognition [27]. The fundamental contributions of both activities to animal health and productivity prompted us to investigate the structure and function of the SPAG11 gene in cattle. Financial and food quality losses resulting from mastitis [28, 29] are driving active investigation of bovine defensins and their regulation in mammary gland. However, defensins in the bovine male reproductive tract, including SPAG11, have received little attention. In this report, the exon structures of the fused bovine SPAG11 gene and of the mosaic transcripts initiated at both A and B promoters are established, including identification of novel exons and transcripts not previously found in primate or rodent. Transcripts are differentially expressed in different tissues from the male reproductive tract of fetal and adult bulls. Further, we demonstrate the presence of SPAG11C and SPAG11D/E proteins in the epithelium of epididymis and testis and SPAG11D in and on epididymal spermatozoa and its function as a potent antibacterial agent.
Testis, epididymis (caput, corpus, and cauda), and vas deferens were collected from adult bulls (Bos taurus) at local abattoirs in Botucatu, São Paulo, Brazil, and Godwin, North Carolina. Tissues were isolated, snap frozen in liquid nitrogen, and stored at –75°C prior to processing. Bovine tissues obtained from Martin's Abattoir and Wholesale Meats (Godwin, NC) were provided under a permit to obtain research specimens issued by the North Carolina Department of Agriculture and Consumer Services.
Full-Length Cloning of Bovine SPAG11
Total RNA from bovine adult caput epididymis was extracted by the method of Chirgwin et al. [30] and poly(A)+ RNA extracted using Oligotex mRNA midi kit (QIAGEN, Valencia, CA), according to manufacturer's instructions. Full-length cloning of SPAG11 cDNAs was achieved using the caput epididymis poly(A+) RNA in two experimental strategies: PCR screening of a directional cDNA library and rapid amplification of 5'- and 3'-cDNA ends (5'and 3'-RACE).
Computational methods (BLAST search at the Genome Sequencing Center at Baylor College of Medicine, http://www.hgsc.bcm.tmc.edu) revealed that the Bos taurus chromosome 27 genome sequence contains the bovine SPAG11 fusion gene (Supplemental Data 1, available online at www.biolreprod.org). Alignment of the bovine sequence with published primate and rodent SPAG11 transcripts enabled preliminary estimation of intron/exon boundaries and the design of the initial primers F, DF, R, and DR used for PCR screening of a cDNA library (Table 1 and Fig. 1). The amplified DNA sequences provided further information for the design of the additional primers F0, UF, EF1, R6, CR used for RT-PCR and 5'RACE strategies (Table 1 and Fig. 1). Introns and exons were then estimated based on nucleotide alignment with human SPAG11 gene (GenBank accession number AC134395) and bovine genomic sequence (GenBank accession number CM000203).
|
|
The directional cDNA library was constructed in 
ZAPII (Stratagene, Cedar Creek, TX) according to the manufacturer's instructions. The primary library was amplified once and aliquots (5 µl) were screened by PCR analysis as follows: phage primers T3 (5'-AATTAACCCTCACTAAAGGG-3') or T7 (5'-GTAATACGACTCACTATAGGGC-3') in combination with one of the SPAG11-specific forward (F or DF) or reverse primers (DR or R) were used to amplify 5'- and 3'-ends of inserted cDNA clones (Table 1). Reactions were performed in a final volume of 25 µl containing 20 mM Tris-HCl pH 8.4, 50 mM KCl, 1.5 mM MgCl2, 0.4 mM dNTPs, 1.5 U Taq polymerase (Invitrogen, San Diego, CA) and 0.4 µmol of each (sense and antisense) specific primer. PCR protocol consisted of an initial cycle of 1 min at 95°C, followed by 35 cycles of 1 min 95°C, 1 min 60°C, and 1 min 72°C, and a final extension of 3 min 72°C. Aliquots of the DNA samples (18 µl) were loaded onto agarose gels (1.8%) containing ethidium bromide. PCR products were visualized with fluorescent illumination and photographed. Amplicons were subcloned into pCRII (Invitrogen). Inserts were sequenced with an ABI PRISM 377 automated sequencer (Applied Biosystems, Foster City, CA) and BigDye Terminator Sequencing kit (Applied Biosystems) at the DNA sequencing facility located at INFAR, Universidade Federal de São Paulo.
To obtain the full-length SPAG11 cDNAs, the Marathon cDNA Amplification Kit (Clontech, Palo Alto, CA) was used following manufacturer's protocol. Briefly, oligo(dT)-primed cDNA was synthesized from 1 µg of bovine testis and caput epididymis poly(A)+ RNA (1 µg). The Marathon cDNA adaptor (10 µmol) was ligated to the double-stranded cDNA using T4 DNA ligase (1 U/µl). Adaptor-ligated cDNA (1 µl) was then diluted in Tricine-EDTA buffer (250 µl), heated at 94°C for 2 min, and stored at –20°C. RACE PCR was performed in a final volume of 25 µl containing 20 mM Tris-HCl pH 8.4, 50 mM KCl, 1.5 mM MgCl2, 0.4 mM dNTPs and Advantage 2 polymerase (Clontech) in the presence of a nested adaptor primer (AP1, 5'- CCATCCTAATACGACTCACTATAGGGC-3', Clontech) and one SPAG11 specific primer for amplification of 5'- and 3'- ends, respectively (primers F or DR, Table 1). PCR conditions were based on manufacturer's recommendations. When cycling was completed, RACE DNA products were gel-purified, subcloned into pCRII vector (TOPO TA kit, Invitrogen), and sequenced, as previously described.
Nucleotide and Protein Sequence Analysis
Nucleotide sequences of the full-length SPAG11 cDNAs were analyzed using the software BioEdit Sequence Alignment Editor [31] and translated using the Expasy translate tool (http://us.expasy.org/tools/dna.html), revealing the full length amino acid sequences. The degree of conservation of the SPAG11 protein isoforms from bull in comparison to other species (primates, rodents, dogs, pigs) were analyzed by means of the alignment editor GeneDoc (http://www.psc.edu/biomed/genedoc/) with identities and similarities shaded to four levels of conservation according to default parameters. A PROSITE scan was performed to identify consensus posttranslational modification sites [32]. Potential O-glycosylation sites were analyzed by means of the CBS Prediction Server (http://www.cbs.dtu.dk/services/NetGlyc) [33].
Using the BioInfoBank MetaServer (http://bioinfo.pl/Meta) [34, 35] the six-cysteine array sequence of bovine SPAG11C (FKVVRCIKGDGKCQKFCNYMEFQLGYCSKKKDACCL) was threaded onto the structure of mouse defensin beta 8 (DEFB8) [36] in file 1E4R.pdb in the protein data bank (http://www.rcsb.org/pdb/). A model of this bovine SPAG11C peptide was built using the Modeler module of the Insight II molecular modeling system from Accelrys Inc. (San Diego, CA, and http://www.accelrys.com/products/insight/). The SPAG11C model figures were created using PyMol (DeLano Scientific, LLC, San Carlos, CA) in the Structural BioInformatics Core Facility (University of North Carolina at Chapel Hill, NC) under the direction of Dr. Brenda Temple. Static figures were labeled and arranged using Adobe Photoshop (Adobe Systems, Inc., San Jose, CA). A video of this bovine SPAG11C peptide model rotating was created using PyMol and assembled using VideoMach (http://www.gromada.com/videomach.html) and is available as Supplemental Data 2 (available online at www.biolreprod.org).
The nucleotide sequences of the exons encoding the six-cysteine arrays of SPAG11C, SPAG11D/E, defensin beta (DEFB) 103 (DEFB103), and DEFB104 orthologs were aligned using the CLUSTALW program. The following GenBank accession numbers were used: SPAG11C: human NM_058203, macaque AY528234, bull DQ838981, pig BX925543, dog DQ012011, mouse AY552530, rat DQ012093; SPAG11D/E: human NM_058201, macaque AY528235, bull DQ838981, pig BK005523, dog DQ012012, mouse NM_153115, rat NM_145087; DEFB103: human NM_018661 (DEFB103A), bull AJ568025, pig AY460575, dog DQ011972; DEFB104: human BC100850 (DEFB104B), bull from the genomic NW_995797, and dog DQ011973. This alignment file is available as Supplemental Data 3 (available online at www.biolreprod.org). This file was used to identify highly conserved, surface-localized residues that may be functionally and structurally important using the ConSeq server (http://conseq.bioinfo.tau.ac.il/) [37]. This alignment was also used for phylogenetic analyses using the Neighbor-Joining method, using the poisson correction and 500 bootstrap replicates in the MEGA molecular evolutionary genetic analysis software package (Version 3.1) [38].
Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) Analyses
RT-PCR studies were conducted using total RNA from adult and fetal tissues. Total RNA from adult bovine tissues (testis, caput and cauda epididymis, vas deferens) was isolated using TRIzol (Gibco-BRL, Gaithersburg, MD). Total RNA from intestine, kidney, liver, adrenal, and testes (from male bulls) and ovary from both adult and fetal (days 90–150 of pregnancy) animals were kindly provided by Dr. José Buratini Junior (Universidade Estadual Paulista, Botucatu, São Paulo, Brazil) and used for tissue distribution studies. The age of each fetus was estimated from the crown-rump length as previously described [39]. Fetal tissues were collected at a local abattoir in Botucatu, São Paulo, Brazil, where animals were slaughtered in accordance with the local legislation and approval of the Universidade Estadual Paulista.
RT-PCR amplification was performed using ThermoScript RT-PCR system for first strand cDNA synthesis (Gibco-BRL, Gaithersburg, MD). Oligo(dT)-primed cDNAs were synthesized from 4 µg of total RNA from bovine tissues for 1 h at 55°C, in a reaction volume of 20 µl. The resulting cDNAs (2 µl) were amplified by PCR in a final volume of 25 µl containing 20 mM Tris-HCl pH 8.4, 50 mM KCl, 1.5 mM MgCl2, 0.4 mM dNTPs, Taq polymerase (Gibco-BRL) and 0.4 µmol of each sense and antisense specific primer chosen based on the SPAG11 gene structure deduced from the cloning experiments reported above (Fig. 2). Each set of primers spanned at least one intron to ensure that PCR products were from cDNA and not genomic DNA. No-template negative control reactions and RT-PCR in the absence of the reverse transcriptase, in this case to assess DNA contamination in the template RNA, were routinely performed. Primer sequences are showed in Table 1. The pairs of primers, spanning at least one intron, were used to amplify the following specific SPAG11 splice variants: primers F0/CR to amplify SPAG11C (
351 bp) and SPAG11U ( 510 bp); primers F0/DR to amplify SPAG11D ( 382 bp), SPAG11V ( 537 bp), and SPAG11W ( 669 bp); primers EF1/DR to amplify SPAG11E ( 314 bp); primers UF/CR to amplify SPAG11U ( 228 bp); and primers UF/R6 to amplify SPAG11V ( 421 bp).
|
Semiquantitative RT-PCR analysis was used to determine the expression of each specific bovine SPAG11 splice variant. For each pair of primers the number of cycles to amplify each cDNA in the linear range was determined under the following PCR conditions: an initial cycle of 1 min at 95°C, followed by 20–35 cycles of 1 min at 95°C, 1 min at 60°C, and 1 min at 72°C, and a final extension of 3 min at 72°C. The transcripts SPAG11D, SPAG11V, and SPAG11W required 35 cycles for PCR linear detection, while SPAG11C, SPAG11E, and SPAG11U required 28 cycles. Amplification of the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase GAPDH ( 200 bp), used as internal control, required 25 cycles for PCR linear detection. Aliquots of the amplified DNA samples (18 µl) were loaded onto agarose gels (1.8%) containing ethidium bromide. PCR products were visualized with fluorescent illumination and photographed. Amplicons were gel purified, subcloned into the pCRII vector (Invitrogen), and their identity verified by automated sequencing.
Peptides were synthesized using a Rainin Symphony multiple peptide synthesizer (Rainin Instruments, Woburn, MA) using fluorenylmethyloxycarbonyl chemistry in the University of North Carolina Program in Molecular Biology Protein Chemistry Facility under the direction of Dr. David G. Klapper, Department of Microbiology and Immunology, UNC-Chapel Hill. Peptides were purified by HPLC. Peptide sequences for SPAG11C (YQIVNSKKSEGQSQEC) and SPAG11E (ERKGDISSDPWNRC) were based on the mouse sequence and the internal cysteines were replaced with serines to minimize disulfide bond-mediated aggregation. Human SPAG11D/E peptide at the C-terminus (CVSNTDEEGKEKPEMDGRSGI) was also used [8]. These antibodies were originally raised for the purpose of analyzing expression in rodents and humans. When sequence comparisons suggested high probability of cross-reactivity with bovine SPAG11D/E and possible cross-reactivity with bovine SPAG11C, the antibodies were tested in immunohistochemical studies with bovine tissue sections, where they showed readily competed immunostaining in epithelial cells (as described below). C-terminal cysteines were conjugated to keyhole limpet hemocyanin. Two rabbits were injected with each antigen peptide at Bethyl Laboratories (Montgomery, TX) and boosted at 2-wk intervals. Blood was taken 10 days after each boost. For each antigen, sera from one rabbit was used. All procedures were conducted in accordance with standard guidelines concerning the care and use of animals in research, teaching, or testing. Bethyl Laboratories is a Registered Research Facility with the United States Department of Agriculture under the Animal Welfare Act and complies daily with all requirements of the Act. Sera were selected based on clarity of localization and competition by the antigen peptide. Antibodies were affinity purified on peptide antigen columns prepared using the Sulfo-Link kit (Pierce, Rockford, IL) according to the manufacturer's recommendations.
Tissues for immunohistochemistry (adult bovine testis, caput, and cauda epididymis) were fixed in Bouin fluid (75 ml saturated picric acid, 5 ml glacial acetic acid, and 25 ml 37% formaldehyde) promptly after excision and embedded in paraffin as previously described [8]. Sections were pretreated for 30 min at 25°C with 5:1 methanol:30% H2O2 to block endogenous peroxidase. Sections were stained using affinity purified antiserum (1–8 µg/ml) raised in rabbit against the human SPAG11D/E or mouse SPAG11C. For the control staining, antibody was preincubated with the respective antigen peptide used for antibody production. A Vectastain Standard ABC kit (avidin-biotin-complex horse radish peroxidase) (Vector Laboratories Inc., Burlingame, CA) was employed to localize immunoreactive SPAG11 using diaminobenzidine as chromogen, resulting in a dark brown reaction product. Sections were counterstained with toluidine blue. Photographs were taken with a SPOT Cooled Color digital imaging system (Diagnostic Instruments, Inc., Sterling Heights, MI) attached to a Zeiss Photomicroscope III. Photographs were prepared using SPOT image processing software. Images were arranged using Photoshop (Adobe Systems Inc., San Jose, CA).
Recombinant Protein Production
Recombinant proteins were prepared as described earlier [8] with modifications as follows. In brief, E. coli ORIGAMI B(DE3) pLacI (Novagen), which facilitated the formation of disulfide bonds in the cytoplasm, was transformed with a modified vector pQE80–1 (EcoRI-SphI from pQE2; TAGzyme His6 tag sequence MKHHHHHHHMHA, Qiagen, Valencia, CA) according to the supplier's instructions. Vectors included cDNA coding for full length bovine SPAG11D (amino acids 22–129) and C-terminal SPAG11E (amino acids 19–80) (Fig. 3). Fusion protein expression was induced with 1 mM isopropyl-1-thio-ß-D-galactoside for 1 h at 37°C. Bacterial lysate (in 0.1 Mol NaH2PO4, 0.01 Mol Tris, 8 Mol Urea pH 8.0) incubated with nickel-nitrilotriacetic acid-agarose (Qiagen, Valencia, CA) was transferred to a column, washed (in 0.1 Mol NaH2PO4, 0.01 Mol Tris, 8 Mol urea pH 6.3), and eluted (in 0.1 Mol NaH2PO4, 0.01 Mol Tris, 8 Mol urea pH 5.9 and pH 4.5) according to the manufacturer's recommendations. Fractions were analyzed on 10%–20% gradient polyacrylamide Tris-Tricine gels and stained with Coomassie blue G250. Fractions containing purified protein were pooled and dialyzed against 10 mM sodium phosphate buffer (pH 7.4) to remove urea. Recombinant protein yields for SPAG11D full length and C-terminal peptide were about 500 µg/g cells.
|
The colony forming units (CFUs) assay was employed to test the antibacterial activity as described earlier [40]. Briefly, overnight cultures of E. coli XL-1 blue (Stratagene, La Jolla, CA) allowed to grow to midlog phase (A600 = 0.4–0.5) were diluted with 10 mM sodium phosphate buffer (pH 7.4). Approximately 2 x 106 CFU/ml of bacteria were incubated at 37°C with 1–10 µmol of either recombinant full length or the C-terminal peptide of bovine SPAG11 D (same amino acid sequence as SPAG11E). Aliquots of the assay mixture removed at 0–180 min after incubation were serially diluted with 10 mM sodium phosphate buffer (pH 7.4) and 100 µl of each was spread on a LB agar plate and incubated at 37°C overnight to allow colony development. The resulting colonies were hand counted and bacterial survival expressed as CFU/ml.
Immunoblotting and Immunofluorescent Staining of Sperm from Bovine Epididymis
Bovine spermatozoa were isolated from the three regions of the epididymis (caput, corpus and cauda) by allowing sperm to emerge from sliced tissue into phosphate buffered saline (PBS) (0.1Mol NaPO4, 0.15 Mol NaCl, pH 7.2). Sperm isolated from each epididymal region were washed twice in PBS and extracted in 8 Mol urea/ 0.1 Mol NaPO4, pH 8.0. Total protein extracts (from approximately 10 million spermatozoa) were then loaded on each lane of a 4%–12% NuPage gel (Invitrogen, San Diego, CA) run in MES buffer according to Invitrogen recommendations. Proteins were transferred in BisTris/Bicine EDTA (according to Invitrogen recommendations) in 20% methanol to 0.45 micron polyvinylidene difluoride (PVDF) membrane (Immobilon-P, Millipore Inc., Billerica, MA). The migration of the SeeBlue Plus2 pre-stained molecular weight standards (Invitrogen) and human recombinant full length SPAG11D (65 ng) were used as references for the calculation of the apparent molecular weight of the bovine SPAG11D protein band. Blots were blocked in 1% casein in TBS-T (50 mM Tris-HCl, 150 mM NaCl, 0.1% Tween 20; pH 7.6). Immunodetection was performed using specific antibody (dilution 1:5000) raised against modified mouse SPAG11E peptide (ERKGDISSDPWNRC). As a specificity control, an equal amount of antibody was pre-incubated with antigen peptide. Antigen antibody complexes were detected on the blot using horse radish peroxidase-conjugated donkey antirabbit IgG (Amersham Pharmacia, Piscataway, NJ) and Pico West Dura enhanced chemiluminescence extended duration substrate (Pierce Biotechnology, Inc., Rockford, IL).
Immunofluorescent staining using affinity purified rabbit antibody to human SPAG11D/E (50 µg/ml) was performed on spermatozoa isolated from bovine caput epididymis. Sperm were washed four times with PBS, fixed in 2% formaldehyde, and spread on glass slides. The negative control was performed using antibody preincubated with the respective peptide antigen. After overnight incubation of sperm slides with primary antibody, slides were washed four times with PBS, blocked with 4% normal goat serum for 15 min, and incubated with Alexa Fluor 488-conjugated goat anti-rabbit IgG (1:2000) (Molecular Probes, Eugene, OR; Invitrogen, San Diego, CA) for 30 min at room temperature. Slides were washed four times in PBS and mounted in ProLong antifade medium (Molecular Probes). Spermatozoon images were taken using a Zeiss Axiophot microscope with a Zeiss Axiocam digital camera.
Cloning of Bovine SPAG11 mRNA Subtypes
Following analysis of the bovine SPAG11 gene by computational methods in a DNA fragment derived from chromosome 27q1.2, additional exons were discovered by analysis of transcripts (Supplemental Data 1, available online at www.biolreprod.org). PCR screening of a caput epididymis directional cDNA library and 5'-RACE studies performed using total RNA from caput and cauda epididymis revealed the complete coding sequence of six bovine SPAG11 splice variants, representing transcripts derived from promoters A and B. These transcripts indicated that the bovine gene is composed of eight exons (Fig. 1) spanning about 20 kb, somewhat larger than the 13 kb human SPAG11 gene (Fig. 2). The nucleotides present at the splice sites at the exon and intron boundaries are shown in Table 2.
|
Bovine exons bE1, bE2, bE4, bE6, and bE7 have nucleotide sequence similarity to human exons E1, E2, E3, E4, and E6, respectively. In addition, the bovine gene recruited novel exons bE3, bE5, and bE8 not reported in other species (Fig. 2A). Based on nucleotide sequence and predicted amino acid similarities to the human transcripts, bovine SPAG11C, SPAG11D, and SPAG11E splice variants were identified (Fig. 2B). SPAG11C and SPAG11D are under the control of promoter A and contain identical signal peptides encoded by exon bE1. SPAG11E was the only transcript detected under the control of promoter B. It contains the nucleotide sequence present in exon bE6, which encodes the signal peptide for this variant. Similar to human and other species, these three SPAG11 variants all contain the 6-cysteine defensin motif in their C-terminal peptides encoded by exons bE4 (SPAG11C) and bE7 (SPAG11D and SPAG11E) (Figs. 2B and 3B). Interestingly, bovine SPAG11C has a longer C-terminal region compared to human and other species (Fig. 3B). Three novel SPAG11 variants, representing bovine-specific transcripts, were identified and named SPAG11U, SPAG11V, and SPAG11W (Fig. 2B). These variants are under control of promoter A and contain an early stop codon present in exon bE3 that leads to a truncated C-terminal isoform (Figs. 2B and 3B).
The multiple alignments and conservation analysis of the predicted amino acids encoded by bovine exons bE1, bE2, bE4, and bE7 with SPAG11 orthologs from primates (human and rhesus monkey), rodents (mouse and rat), dog and pig revealed different degrees of divergence in different exons (Fig. 3). Amino acid residues in the signal peptide encoded by exon bE1 were less conserved than those encoded by bE6 compared to other species (35%–45% versus 72%–83% amino acid identity, respectively) (Fig. 3A). Bovine exon bE2 codes for 40 amino acids present in the structure of all bovine SPAG11 isoforms under the control of promoter A. This amino acid sequence is more distant from primates (
35% identical) and rodents (36%–42% identical) than dog (50%) and pig (60%) orthologs (Fig. 3B) and shows no homology with any other known proteins. Differences in the degree of evolutionary conservation were also observed in the bovine ß-defensin-like exons bE4 and bE7 (Fig. 3B). The amino acid sequence deduced from the bovine exon bE4 shows 30%–38% similarity, while bovine exon bE7 contains 54%–74% amino acid sequence similarity with the different orthologs. Among all species shown in Figure 3B, the highest levels of amino acid identity and similarity occurred between bovine bE7 (in the SPAG11D and E isoforms) and primate orthologs (74%), while amino acids encoded by bovine bE4 (in the SPAG11C isoform) presented 38% identity and similarity to pig orthologs.
A PROSITE database search for functional domains identified several consensus posttranslational modification sites in bovine SPAG11 isoforms (Fig. 3). A potential N-glycosylation site (NRSH) is present in the N-terminal peptide of isoforms SPAG11C, SPAG11D, SPAG11U, SPAG11V, and SPAG11W. Two protein kinase C phosphorylation sites were identified in isoforms SPAG11C (TNR and SKK), SPAG11D (TNR and TCR) and one site was detected in SPAG11E (TCR). One casein kinase II phosphorylation site was also present in isoforms SPAG11D and SPAG11E (SEEE). Myristoylation sites were detected in isoforms SPAG11C and SPAG11D (GTNRSH). Although not observed in bovine SPAG11C, potential O-glycosylation on the C-terminal S and T amino acids, which are conserved in bovine, human, and rhesus monkey SPAG11D and SPAG11E, were identified.
A model of the six-cysteine array of the bovine SPAG11C isoform shows a short alpha helix and the first and third of the three ß strands typical of a ß-defensin structure (Fig. 4A). The bovine region where the second ß strand would be models as a loop structure. Conserved exposed amino acids identified by ConSeq are indicated in red in Figures 4B and 4C. The model also shows the basic amino acids H, K, and R (blue in Fig. 4D) mainly located on one side of the protein structure.
|
To better understand the evolution of the bovine SPAG11 sequences and their relationships with primate (human and macaque), rodent (mouse and rat), dog, and pig orthologs, a phylogenetic tree was constructed using only the exons encoding ß-defensin-like sequences (bE4, bE7, and orthologs) (Fig. 5). Because ß-defensin genes have proliferated primarily by gene duplication [41], we hypothesize that DEFB104B, the gene next to SPAG11 A component (SPAG11C isoform), and DEFB103A, the gene adjacent to the SPAG11 B component (SPAG11E isoform) in humans (probably similarly arranged in cattle), evolved by gene duplication from the common ancestor of SPAG11. The results show the SPAG11 exons and the two defensins forming three major clades. The high bootstrap values for many branches in the tree give confidence in the tree structure. The tree supports the conclusion that SPAG11E diverged early from the common ancestor that gave rise also to a progenitor of SPAG11C and an early ß-defensin that duplicated and evolved to form DEFB103 and DEFB104. The bull and later pig SPAG11C diverged early from the other orthologs, whereas bull SPAG11E has maintained closer relatedness to its orthologs, especially primates.
|
Expression of SPAG11 mRNA Splice Variants in Fetal and Adult Bovine Tissues
The distribution of SPAG11 splice variants and their relative expression were analyzed in adult bovine testis, epididymis (caput, corpus, and cauda), and vas deferens (Fig. 6). All six alternative splicing variants were differentially identified in the different tissues from the male reproductive tract. Variant transcripts SPAG11E and SPAG11U were abundant in all tissues analyzed, while transcripts SPAG11V and SPAG11W were mainly amplified in testis. Although no SPAG11D mRNA was detected in vas deferens, this variant transcript was readily observed in testis and caput epididymis, with lower abundance in corpus and cauda epididymis. SPAG11C mRNA was detected in all male reproductive tissues, although in lower abundance in the testis (Fig. 6).
|
The distribution of SPAG11 splice variants in fetal and adult bovine testis was compared to nonreproductive tissues, including intestine, kidney, liver, adrenal, and ovary (Fig. 7). Expression levels of each SPAG11 transcript depended on the age and tissue analyzed. The greatest differences between fetal and adult were seen in levels of transcripts SPAG11E, SPAG11U, SPAG11V, and SPAG11W (Fig. 7, A–C), which were marginally detected in fetal testis but abundant in adult testis. Interestingly, levels of SPAG11C mRNA were only detected in fetal kidney, adrenal, and ovary, but no prominent amplification of this transcript was observed in the adult tissues other than testis (Fig. 7A). Similar transcript levels of SPAG11D were detected in all fetal tissues examined, although this transcript variant was not detected in intestine or adrenal in the adult animal (Fig. 7B). SPAG11U mRNA was detected in the fetal kidney but not the adult kidney (Fig. 7A).
|
Immunohistochemical Detection of SPAG11 Isoforms
Consistent with the transcript amplification analysis, SPAG11C and SPAG11D/E were specifically immunodetected in the epithelial cells and in the tubule lumen of bovine caput epididymis and testis (Fig. 8). SPAG11D/E immunoreactivity was detected in the epithelium throughout the epididymis, but was most abundant in the distal efferent ducts (data not shown) and proximal caput (Fig. 8, upper panels A and B). SPAG11D/E was most prominent in the apical region of the epithelial cells and was detected in the lumen together with the sperm. In the testis SPAG11D/E was immunolocalized in late stage spermatids and in Sertoli cells (Fig. 8, lower panels A and B). In the caput epididymis, SPAG11C was expressed in epithelial cells and was concentrated among the microvilli (Fig. 8, upper panels C and D). SPAG11C positive staining was most prominent in late spermatids and also present in Sertoli cells (Fig. 8, lower panels C and D).
|
SPAG11D Protein is Associated with Sperm Obtained from Bovine Caput Epididymis
Western blot analysis indicated SPAG11D immunoreactivity in protein extracts obtained from 10 million spermatozoa isolated from caput, corpus, and cauda epididymis (Fig. 9A, lanes 1–3) migrating with an apparent size of approximately 12 kDa, which corresponds closely to the migration of recombinant human SPAG11D and to the calculated size of bovine SPAG11D of 12.5 kDa. Comparison of the band intensities observed with 5 pmol (65 ng) of recombinant human full-length SPAG11D (Fig. 9A, lane 4) to the band observed with extracts obtained from 10 million sperm isolated from bovine caput epididymis allowed us to estimate that approximately 0.75 pmol (or 10 ng) of SPAG11D is associated with one million caput sperm. The presence of SPAG11D in or on spermatozoa isolated from bovine caput epididymis was demonstrated further by immunofluorescent staining (Fig. 9B). It was observed that SPAG11D was localized to the entire sperm tail including the middle piece and neck. Antibody pre-incubated with antigen peptide failed to produce specific immunostaining (data not shown).
|
Bovine SPAG11D and SPAG11E Antibacterial Activity
To determine if antibacterial activity is conserved as is the sequence homology with primate orthologs and other defensins, full length bovine SPAG11D and SPAG11E peptide were tested for activity against E. coli. (Fig. 10). The full-length SPAG11D protein was strongly antibacterial and the SPAG11E C-terminal peptide that contains the defensin motif was somewhat less potent. Thus, although the N-terminal peptide contains no defensin-like functional motif, it either contains intrinsic antibacterial activity as demonstrated for the human and monkey orthologous peptides [24], or its presence in the SPAG11 molecule enhances the antibacterial activity of the C-terminal peptide.
|
In the present study, molecular analysis revealed that the Bos taurus SPAG11 gene spans a DNA fragment of approximately 20 kb and expresses eight exons. Chromosomal region 27q1.2 is the likely site of the bovine SPAG11 locus because in this region and human 8p23 are established positions of marker genes malignant fibrous histiocytoma amplified sequence 1 (MFHAS1) and defensin beta 1 (DEFB1) that flank the human SPAG11 gene [42]. This conclusion is supported by cytogenetic mapping of this DEFB gene cluster to bovine chromosome 27q [43] and the location of DEFB1 near the centromere of chromosome 27 in Bos taurus genome build 3.1.
The entire coding sequences were determined for six bovine SPAG11 mRNA variants whose expression is regulated by promoter choice and alternative splicing mechanisms. Five of these variants were initiated at promoter A (SPAG11C, SPAG11D, SPAG11U, SPAG11V, and SPAG11W), while one was initiated at promoter B (SPAG11E). Initially discovered in human [8], the ability of the SPAG11 gene to produce both fusion and independent transcripts is here shown to be conserved in cattle, which diverged from the human line 92 million years ago [44]. Conservation of this transcriptional versatility suggests a fundamental fitness advantage for SPAG11 function. One advantage, as reported for orthologs in other species [10, 17], is the capacity of bovine SPAG11C, SPAG11D, and SPAG11E mRNAs for differential expression in different tissues where each may have a somewhat different function.
The linkage of the two SPAG11 A and B component defensin genes is unique in the defensin family. None of the other more than 40 defensin genes is known to be transcriptionally fused to its defensin neighbor or any other neighboring gene. What is it about these two ß-defensin genes, the SPAG11 A and B components, that makes them work well together as a fusion in genomes as different as primate and bovine? Properties that predispose these particular defensins to linkage include physical proximity on the chromosome and the same direction of transcription. The SPAG11 component genes are about 10 kb apart compared with 8.5 kb median intergenic distance for known fused genes and compared with 48 kb for the median distance between adjacent human genes in general [45]. Fusion is not simply promoted by proximity since several pairs of defensins are closer and are transcribed in the same direction. For example, in the human genome, DEFB104B and DEFB106B are 7.5 kb apart and DEFB104B and SPAG11 are 6 kb apart, yet fusion transcripts have not been reported. DEFB105B and DEFB106B are closer still (1.3 kb apart), but transcribed in opposite directions. Benefits of gene fusion include bringing domains into the same molecule that otherwise could interact as independent proteins. As a well-analyzed example of such benefits, eukaryotic nitric oxide synthase genes evolved from prokaryotic predecessors by a series of gene fusion events producing a modular enzyme [46]. The alignment of the molecular surfaces of the different domains is a central feature of structure, catalysis, and control in these enzymes. There may be cooperative advantages in having different functional modules covalently associated in the same polypeptide chain. Such benefits may be operational in SPAG11 protein isoforms.
The SPAG11 A component may have a semifunctional transcriptional termination site that, when active, preempts continuation to the B component. Models of transcription termination indicate that both cis-acting sequence elements and trans-acting termination factors that belong both to the transcriptional and splicing machineries act together to generate a 3' terminus [47, 48]. Since splicing occurs cotranscriptionally, a splice site encountered just prior to a weak polyadenylation signal could deflect termination [48]. Promoters and the transcription activators they attract can also influence splicing [49–51]. Thus, species differences in alternative splicing may reflect promoter sequence differences. Splicing mechanisms also may be dependent on optimal 5' and 3' splice sites [52]. The present study shows that a different pattern of nucleotide sequence is involved in each splice site junction in bovine SPAG11 variant transcripts, and their contribution to splicing regulation remains to be investigated.
The proximity of the two defensin genes DEFB103A and DEFB104B immediately adjacent to human SPAG11 is consistent with evolution of these three genes from a relatively recent common ancestor. Evidence for part of this concept was presented in a recent phylogenetic analysis [13], which shows SPAG11C and DEFB103A evolved from a common ancestor that evolved from the same ancestor as SPAG11E. Similar trends are evident in the analysis of bovine SPAG11 sequences compared with primate, rodent, porcine, and canine orthologs in the present study. This phylogenetic tree supports the conclusions that SPAG11C, DEFB103, and DEFB104 descended from a common ancestor from which SPAG11E diverged slightly earlier. Differential evolution of exons is seen in the bovine exon 4 (SPAG11C), which diverged early from the other orthologs, whereas bovine exon 7 (SPAG11E) has maintained closer relatedness with its orthologs. The divergence of these genes may be related to the fact that the chromosomal region where SPAG11 is located in human, dog, mouse, and presumably cattle has the highest divergence levels seen across all autosomes [53].
Overarching concepts of gene function and the elements responsible for those functions are developed through cross-species comparisons that reveal how a gene has responded to evolutionary pressures in the varied environments of different genomes. Apparently propelled by both immunological and reproductive demands, SPAG11 isoforms, though diverse, have maintained a few highly conserved residues as seen in the multiple species comparisons with bovine sequences. These conserved exposed amino acids identified by ConSeq may have structural importance, as do the cysteines that form disulfide bonds providing stability. Since these specific residues are thought to mediate defensin dimerization [25] it can be speculated that the conserved SPAG11C and SPAG11D/E sites may be involved in homo- or heteromolecular recognition in host defense and reproductive pathways. In fact, dimer formation could be important in bacterial membrane disruption activities, as suggested for certain conserved positions in human DEFB103A [54]. Alternatively, the conserved exposed amino acids may be conserved because of key functional roles, such as interacting with receptors or other binding partners. It is interesting to note that potential protein kinase C and casein II phosphorylation sites are present in bovine SPAG11 protein isoforms. Phosphorylation alters the function of the protein by allosteric modulation and by affecting its possible interactions with other molecules and possible repertoire of dimer formation [55]. Experimental evidence will be necessary to determine functional significance of the phosphorylation sites and how they might interfere in possible bovine SPAG11 homo- or heteromolecular recognition events.
Differential expression of bovine SPAG11 variant transcripts in tissues from adult bulls shows that bovine SPAG11C, SPAG11E, and SPAG11U transcripts are abundantly expressed in the male reproductive tract, similarly to primate SPAG11 variants and other defensins [10, 41]. In the nonreproductive tissues tested, SPAG11E mRNA is detected only in kidney, while lower abundance of SPAG11C and SPAG11U transcripts is detected in ovary and adrenal gland. In addition, the variant transcripts SPAG11W and SPAG11V are detected only in testis. In contrast, bovine SPAG11D mRNA is expressed in the caput epididymis and testis and in nonreproductive tissues, suggesting a broader molecular function. It is important to point out that all six bovine SPAG11 variant transcripts are also differentially expressed when reproductive and nonreproductive tissues from both fetal and adult animals are compared. Taken together, these results suggest highly specific mRNA splicing mechanisms governing bovine SPAG11 transcription rates and or mRNA stability during pre- and postnatal development. It is known that SPAG11 gene promoter selection by transcription complexes exhibits species- and tissue-specific regulation [8, 10, 14]. SPAG11 gene expression is also known to be positively regulated by androgens in primate and rat [8, 17]. Besides androgens, additional tissue-specific factors, hormones and/or extracellular signals differentially present in fetal and adult tissues likely influence the regulation of SPAG11 mRNAs in bovines. Other immune system proteins, such as the major histocompatability Ia group proteins, have established involvement in organ development [56]. Uncovering the identity of these factors involved in regulating SPAG11 expression may provide new insights to the role of SPAG11 protein isoforms in the development of mammalian innate immunity and reproductive functions.
The modular structure of most SPAG11 protein isoforms consists of an N-terminal peptide coupled to a C-terminal peptide encoded by combinations of different exons. Novel bovine-specific transcripts SPAG11U, SPAG11V, and SPAG11W encode a protein with only two amino acids added to the N-terminal region that is common also to SPAG11C, SPAG11D, and other variants. The SPAG11U, SPAG11V, and SPAG11W proteins may also function in innate immunity in spite of the absence of a defensin-like motif, since the orthologous N-terminal domain in primates was shown to have antibacterial activity [17, 23]. The N-terminal domain also appeared to be responsible for most of the antibacterial activity of the full length macaque SPAG11C, SPAG11K, and SPAG11L [17]. The functions of the alternatively spliced 3' untranslated regions of the SPAG11U, SPAG11V, and SPAG11W transcripts are not known but these ends may confer different levels of mRNA stability [57]. It should be pointed out that, unlike classical cationic defensins, the C-terminal peptides of human and macaque SPAG11C are anionic and lack antibacterial activity [17, 23, 58]. However, the orthologous bovine SPAG11C peptide is cationic and thus seems likely to have antibacterial activity similar to that of SPAG11D shown in this report.
Similar to SPAG11D/E expression in primates [8], bovine SPAG11D/E immunoreactivity was detected in the epithelium throughout the epididymis, most abundantly in the distal efferent ducts and proximal caput. SPAG11D/E immunolocalization in the testis and epididymis indicate that it could act as an antibacterial protein in both organs. Further, the detection of SPAG11C and SPAG11D/E in late stage of spermatids and in Sertoli cells suggests a possible role in spermatogenesis as recently proposed for rat SPAG11C [17] and/or a role in antimicrobial protection during sperm passage through the male and female tracts. SPAG11D associated with the tails of spermatozoa may affect sperm motility consistent with the report that rat SPAG11E promotes motility in caput sperm [26]. The apparent size for bovine SPAG11D detected by Western blot analysis (approximately 12.1 kDa) suggests that the immunodetected protein associated with sperm is the whole mature bovine SPAG11D and not SPAG11E (7.1 kDa) and not an N-terminally proteolytically cleaved SPAG11D that corresponds to the major form reported in human [21]. These results may suggest that full-length SPAG11D protein should be functionally relevant in the bovine. Alternatively, the immunodetected protein could be SPAG11E or N-terminally processed SPAG11D enlarged by N- or O-glycosylation of key residues.
Human SPAG11 peptides, due to the cleavage of the furin-like preprotein convertase motif (AVKR) located in the common region encoded by exon 2, are reported in the epididymal fluid, ejaculate, and on the sperm surface. When tested in vitro, these proteolytic processed peptides present antimicrobial activity against Escherichia coli [21]. A furin-like preprotein convertase motif (RVKR) has been also shown in rat SPAG11 isoforms [17]. Protein sequence analysis indicates that the common region encoded by exon bE2 of bovine SPAG11 isoforms contains the paired basic residues KR, which are conserved in SPAG11 isoforms of all species analysed (Fig. 3). Taking into consideration that furin-like prohormone convertases specifically cleave after paired basic residues (especially KR and RR) [59], the data from this study may contribute to a new level of understanding of the cleavage mechanism of SPAG11 isoforms, suggesting that in all species the protein processing through the recognition of the KR paired amino acids is a possibility. However, influence of different neighboring amino acids on the cleavage efficiency cannot be ruled out. Further experiments will be needed to determine the possible proteolytic processing of the SPAG11 isoforms and their existence in the epididymis and other tissues from the male reproductive tract of bovines and other species.
In this report, we establish the bovine SPAG11 gene maintains features observed in primate and rodent orthologs [10, 13, 16–18]. These features include conserved chromosomal location within a cluster of ß-defensin genes, conserved fusion gene structure producing multiple transcripts, species-specific exons, dominant expression in the male reproductive tract, presence of the protein associated with spermatozoa, and antibacterial activity. The forces that determine SPAG11 structure and function in these mammalian lineages may be related to the evolution of SPAG11 isoforms to perform both immune and reproductive functions. Not unique to SPAG11 nor to defensins, the highly specific molecular recognition properties characteristic of immune protein domains are applied in multiple systems where discriminative protein association is crucial. These applications include organ development, cell adhesion, and signaling as well as neuronal differentiation and mature function [52, 60, 61].
Our study defining the gene organization and protein sequence of bovine SPAG11 isoforms contributes to the understanding of the biology of defensins and their functional roles within and beyond the male reproductive tract.
ACKNOWLEDGMENTS
We thank Dr. José Buratini Junior (Universidade Estadual Paulista, Botucatu, São Paulo, Brazil) for his helpful advice.
FOOTNOTES
3Current address: Department of Biochemistry and Molecular Biology, Pondicherry University, Podicherry, 605014, India. 
1Supported partially by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP, grant number 05/55738-8), Brazil, and by the Fogarty International Center Program for Training and Research in Population and Health, USA D43TW/HD00627 (subcontract UNIFESP/UNC 5-53284). Support was provided by the CICCR Program of the Contraceptive Research and Development Program, Eastern Virginia Medical School. The views expressed by the authors do not necessarily reflect the views of CONRAD or CICCR. Also supported by grants from the Andrew W. Mellon Foundation and the National Institutes of Health (R37-HD04466, through cooperative agreement U54-HD35041, as part of the Specialized Cooperative Centers Program in Reproduction Research).
Correspondence: 2Maria Christina W. Avellar, Section of Experimental Endocrinology, Department of Pharmacology, Universidade Federal de São Paulo, Escola Paulista de Medicina, Rua 03 de maio 100, INFAR, Vila Clementino, São Paulo, SP, Brazil, 04044-020. FAX: 55 11 5576 4448; e-mail: avellar{at}farm.epm.br
Received: 20 December 2006.
First decision: 7 January 2007.
Accepted: 20 February 2007.
REFERENCES
