| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Testis |
The Laboratories for Reproductive Biology4
Department of Pediatrics,5 University of North Carolina School of Medicine, Chapel Hill, North Carolina 27599
State Key Laboratory of Molecular Biology6 , Institute of Biochemistry and Cell Biology, Shanghai Institute for Biological Science, Chinese Academy of Sciences, Shanghai, 200031 China
Curriculum in Genetics and Molecular Biology,7 University of North Carolina School of Medicine, Chapel Hill, North Carolina 27599
Laboratory of Pulmonary Pathobiology,8 Cell Biology Section, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709
Department of Cell and Developmental Biology,9 University of North Carolina School of Medicine, Chapel Hill, North Carolina 27599
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
80 kb) were isolated and sequenced. The 3'-UTRs of the two transcripts are quite distinct. The 7.4 kb transcript, which appears earlier in spermatogenesis, has a very long 3'-UTR of 4451 nucleotides. In contrast, the 2.1 kb transcript, which is expressed predominantly during the haploid phase of spermatogenesis, has a 3'-UTR that is only 202 nucleotides in length. Additional analyses indicate that both transcripts share the same coding region and are associated with polysomes. A single GCNF protein band was detected in testis extracts by Western blotting with a specific antiserum. Immunohistochemical analysis showed that GCNF is localized in the nuclei of pachytene spermatocytes and round spermatids. GCNF is first detectable in early pachytene spermatocytes (stage II) and is continuously expressed until spermatids begin to elongate in stage IX. Although GCNF is generally distributed throughout the nucleus, it is particularly prominent in heterochromatic regions at some stages and in condensed chromosomes undergoing the meiotic divisions. This expression profile suggests that GCNF plays a role in transcriptional regulation during meiosis and the early haploid phase of spermatogenesis.
GCNF, gene regulation, nuclear orphan receptor, spermatogenesis, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
An orphan receptor without an identified ligand, GCNF has several distinctive features [11]. It has been classified as the first member of the 6A subfamily of nuclear receptors [12]. Although GCNF has the modular structure characteristic of nuclear receptors, it lacks the conserved transcriptional activation function 2 sequence in the ligand binding domain [13]. GCNF binds as a homodimer to DNA, either to a direct repeat hormone response element without additional nucleotides between the half-sites (DR0) or to extended half-sites [1, 14, 15]. It does not form heterodimers with RXR, a frequent heterodimerization partner of nuclear receptors [13]. In vitro studies suggested that GCNF acts as a transcriptional repressor [6, 16] and may inhibit transcriptional activation mediated by other nuclear receptors [17]. One target of GCNF repression appears to be Oct4, a transcription factor expressed during early germ cell development [18]. Protamine 1 and protamine 2, two genes expressed only during the haploid phase of spermatogenesis, have also been identified as potential targets for GCNF regulation [15, 19].
Another distinctive feature of the Gcnf gene is that two transcripts have been identified that differ in size by 5 kilobase (kb). Only the larger 7.4 kb Gcnf transcript has been identified in mouse embryos [7], embryonal carcinoma cell lines [20], and in some somatic tissues at significantly lower levels [1, 21]. In contrast, spermatogenic cells express both 7.4 and 2.1 kb Gcnf transcripts [2, 4, 5]. The 7.4 kb transcript is expressed earlier during testicular development and is the predominant Gcnf mRNA in pachytene spermatocytes [4]. The smaller Gcnf transcript is expressed predominantly in haploid round spermatids [4, 5].
Although two Gcnf transcripts have been identified in mouse [1, 2, 4, 5], rat [4, 5], and human testes [3, 21], little is known about the structural or functional differences between these transcripts. The aims of this study were to characterize the structure and sequence of Gcnf and its distinct transcripts and to determine the expression and distribution of the GCNF protein during mouse spermatogenesis. Our detailed analyses provide the first evidence that the majority of the 2.1 and 7.4 kb transcripts share the same coding region, with differences confined to the 5'- and 3'-untranslated regions (UTRs). Polysomal analysis suggests that both Gcnf mRNAs are translated. Consistent with its proposed function as a transcriptional repressor, the GCNF protein is particularly abundant in heterochromatic regions of pachytene spermatocyte and round spermatid nuclei and in condensed chromosomes during the meiotic divisions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Testes were isolated from CD-1 mice obtained from Charles River Laboratories (Raleigh, NC). All procedures involving animals were approved in advance by the Institutional Animal Care and Use Committee and were conducted in accordance with the National Research Council Guide for Care and Use of Laboratory Animals. Spermatogenic cells were prepared by sequential enzymatic dissociation of adult testes, and pachytene spermatocytes and round spermatids were isolated by unit gravity sedimentation [22]. Purities of pachytene spermatocytes and round spermatids (steps 1–8) exceeded 90%.
Isolation and Sequencing of Gcnf Genomic DNA
Genomic DNA was isolated from mouse liver (129/OlaHsd strain) as previously described [23]. In addition, a P1 plasmid containing Gcnf genomic DNA from the same mouse strain (150 kb) was obtained from Genome Systems (St. Louis, MO). P1 DNA was prepared according to the instructions provided by the company.
Genomic DNA fragments containing Gcnf coding sequences were generated by subcloning the Gcnf P1 plasmid or by PCR using isolated genomic DNA as a template. Sequences extending beyond the known Gcnf cDNA were generated by genome walking analysis (Clontech, Palo Alto, CA) according to the manufacturer's instructions and confirmed by genomic DNA library screening. DNA sequences were determined using an ABI 3100 sequencer in the DNA Sequencing Facility (University of North Carolina at Chapel Hill) and were assembled and analyzed using the Wisconsin Package Version 10.2, Genetics Computer Group (GCG, Madison, WI). The resulting Gcnf genomic sequences were submitted to the GenBank database (accession numbers AF390897, AF390898, AF390899, AF390900).
Rapid Amplification of cDNA Ends (5'-RACE and 3'-RACE)
5'- and 3'-RACE were carried out using mouse testis Marathon-Ready cDNA (Clontech). Primer sequences were 5'-GAA GGT CAG CAA CGA GGA GTA CG-3' for 3'-RACE and 5'-GTA GTC GTC GTC CTG GTC TTC CAC-3' for 5'-RACE. The 5'-RACE system (version 2.0) from Life Technologies was also used to verify 5'-RACE results. Southern blot analysis was used to confirm that PCR products were derived from Gcnf cDNA. PCR products were excised from agarose gels and purified using Whatman Microspin columns (Whatman, Clifton, NJ). Purified DNA fragments were ligated into the PCRII vector using the TA cloning system (Invitrogen, Carlsbad, CA) and were transformed in E. coli XL1-Blue competent cells using standard molecular biological techniques. The clones were sequenced in the DNA Sequencing Facility (University of North Carolina at Chapel Hill). The derived Gcnf cDNA sequence was submitted to the GenBank database (accession number AF390896).
Genomic Library Screening
A 129 SVJ mouse genomic library in Lambda FIX II (Stratagene, La Jolla, CA) was screened as previously described [24]. Briefly, the library was plated, transferred to nylon filters, and screened using a 405 base pair (bp) probe located before the DNA-binding domain (DBD) of mouse Gcnf (nt 112–525 of AF390896). Positive clones from the first screening of 1 x 106 plaques were plated and subjected to second screening using the same probe. Candidate clones were plaque-purified and confirmed by Southern blotting using a second Gcnf probe (nt 735–1084 of AF390897).
Southern Blot Analysis
Genomic (10 µg/lane) and P1 plasmid DNA (300 ng/lane) were digested with restriction endonucleases (BamHI, HindIII, XbaI, KpnI, and NcoI, Life Technologies, Rockville, MD), and fragments were separated on 0.7% agarose gels. The separated DNA fragments were transferred to Hybond N+ nylon membranes (Amersham Pharmacia Biotech, Piscataway, NJ) using the downward alkaline capillary transfer method [25]. Blots were hybridized with a 238 bp [32P]-labeled probe generated from the Gcnf ligand-binding domain (nt 1001–1238 of AF390896) using ULTRAhyb (Ambion, Austin, TX) according to the manufacturer's instructions. Washed blots were exposed to x-ray film or to a PhosphorImager screen followed by analysis with a STORM869 PhosphorImager (Molecular Dynamics, Sunnyville, CA).
RNA Isolation and Northern Blot Analysis
Total RNA was isolated from mouse testis, pachytene spermatocytes, or round spermatids using the acid-phenol guanidine isothiocyanate method [26]. RNA samples were denatured, electrophoresed on 1.2% formamide agarose gels, and transferred to Hybond-N+ nylon membranes (Amersham Pharmacia Biotech). Transcript sizes were determined by comparisons with markers designed for accurate RNA size determination on Northern blots (Millennium Markers, Ambion). RNA was fixed to the membrane using a Stratalinker UV cross-linker (Stratagene). Blots were hybridized with radioactive probes at 65°C in Church buffer as described previously [27]. In some experiments ULTRAhyb (Ambion) was used for hybridizations according to the manufacturer's instructions. Blots were exposed either to x-ray film at -70°C or to a PhosphorImager screen for analysis with a STORM869 PhosphorImager (Molecular Dynamics).
Specific Amplification of Each Gcnf Transcript
Total RNA (3.0 µg) isolated from testes of adult CD-1 mice was used for reverse transcription (RT) using the ThermoScript RT-PCR System according to the manufacturer's instructions (Life Technologies). Primer L was generated from the 3'-UTR region that is specific to the 7.4 kb Gcnf transcript (nt 2770–2794 of AF390896) and was used for RT. The derived cDNA was used for PCR reactions with primer L and a forward primer generated from the first exon (nt 121–146 of AF390896). A primer specific for the 2.1 kb Gcnf (GSP-T18) was generated to anneal to the last eight nucleotides of the 3'-UTR and the initial segment of the poly (A) tail (illustrated in Fig. 5). Following RT with this primer, PCR amplification was performed with the same forward primer used to amplify sequences from the 7.4 kb transcript and a reverse primer derived from exon 10 (nt 1744–1772 of AF390896). The cDNA products from each RT-PCR reaction were cloned into the TA vector (Invitrogen) for sequence analysis.
|
Polysome Analysis
Polysome gradients were performed as described previously [28–30]. Briefly, isolated spermatogenic cells (3.7 x 107 pachytene spermatocytes, 1.7 x 108 round spermatids) were homogenized with a Dounce homogenizer on ice in 1.0 ml lysis buffer (100 mM NaCl, 1.5 mM MgCl2, and 20 mM Hepes, pH 7.6). The nuclei and mitochondria were pelleted by centrifugation for 5 min at 12 000 x g. Supernatants were loaded onto linear gradients of 15%–50% sucrose (w/w) in lysis buffer over a 0.5-ml 70% sucrose cushion, and were centrifuged for 3 h at 39 000 rpm in a Beckman SW-40 rotor. The gradients were fractionated into ten 1.2-ml fractions, and the RNA in each fraction was precipitated with 0.3 M sodium acetate and isopropanol. RNAs were dissolved in DEPC-treated water, purified with the RNeasy mini kit (Qiagen, Valencia, CA) and subjected to Northern blot analysis (20% of the RNA isolated from each fraction). Control gradients, with 20 mM EDTA replacing MgCl2 in the buffer, were centrifuged and analyzed in parallel to the polysome gradients using 2% of the total RNA isolated from each fraction. After hybridization with radioactive probes specific for Gcnf (nt 1001–1238 of AF390896) or phosphoglycerate kinase-2 (Pgk2, nt 1010–1350 of NM_031190), Northern blots were exposed to Kodak BioMax MS (for pachytene spermatocytes) or X-Omat AR x-ray films (for round spermatids) at -70°C overnight. The Pgk2 probe binds to a Pgk2 region that shares 82% identity with the Pgk1 mRNA, suggesting that this probe may recognize transcripts for both isozymes.
Western Blot Analysis
Testes from adult mice were decapsulated and homogenized in lysis buffer (2% SDS, 100 mM DTT, 60 mM Tris pH 6.8, 10% glycerol). Samples were denatured for 10 min at 85°C, centrifuged at 10 000 x g for 10 min, and protein concentrations were determined with the bicinchoninic acid assay (Pierce Chemical Company, Rockford, IL). Testis proteins (30–50 µg/lane) and molecular weight markers (BenchMark Protein Ladder, Life Technologies) were separated on 10% SDS-polyacrylamide gels and transferred to PVDF membranes (Millipore Corporation, Bedford, MA) for 90 min at 100 mA using a semidry blotter.
Western blot analysis was performed with an antiserum raised against a unique peptide sequence in GCNF (DSDHSSPGNRASESNQPSC). Previous studies demonstrated that this antiserum specifically immunoprecipitates GCNF and causes a supershift of GCNF-nucleotide complexes in electrophoretic mobility shift assays [15]. Blots were blocked with 5% nonfat dry milk and incubated with this GCNF antiserum (diluted 1:1200) followed by an affinity-purified secondary antibody conjugated to horseradish peroxidase (1:30 000, goat anti-rabbit IgG; Kierkegaard and Perry Laboratories, Gaitherburg, MD). Control blots were incubated under identical conditions, except that the primary antibody was preincubated with the GCNF peptide (60 µg/ml) used to raise this antiserum. The blots were washed after each incubation and immunoreactivity was detected with a chemiluminescent substrate for peroxidase (SuperSignal West Pico substrate, Pierce, Rockford, IL).
Immunohistochemistry
GCNF was localized on paraffin sections of Bouin fixed mouse testes using the avidin-biotin immunoperoxidase method (Vectastain ABC kit; Vector Laboratories, Burlingame, CA) as described previously [31]. Prior to immunostaining, the sections were immersed in 1.0 mM EDTA in phosphate-buffered saline (pH 8.0) and microwaved (750 W, three 5 min cycles) for antigen retrieval [32]. The same GCNF antibody used for Western blots was used at concentrations of 1:3000 to 1:4000 for immunohistochemistry. For control sections, the GCNF antiserum was preincubated with the immunogenic peptide (60 µg/ml). Immune complexes were detected using 3'3'-diaminobenzidine tetrahydrochloride (Aldrich Chemical Company, Milwaukee, WI) as the chromogen. Slides were counterstained with Gill hematoxylin (Fisher Scientific, Pittsburgh, PA) and photographed using a Spot RT Slider digital camera (Diagnostic Instruments, Sterling Heights, MD).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Gcnf encodes two transcripts that differ in length by more than 5 kb. The initial report that Gcnf is a single-copy gene [1] was based on Southern blot analysis using a probe that spanned the cloned 3'-UTR, apparently including an extensive region that is unique to the large transcript [4]. To determine if both Gcnf mRNAs are encoded by one gene, we digested genomic DNA with five restriction endonucleases and performed Southern blot analysis using probes common to both transcripts. A 238-bp probe derived from the ligand-binding domain (nt 1001–1238 of AF390896) hybridized to a single DNA fragment in each lane of the Southern blots (Fig. 1, A and B). This probe also hybridized to both 2.1 and 7.4 kb Gcnf transcripts on Northern blots (Fig. 1C). Since sequence data was derived from both P1 plasmid DNA containing Gcnf genomic sequence and from mouse liver DNA, we confirmed that the restriction patterns were identical for both sources of Gcnf genomic DNA (compare Fig. 1, A and B). We also probed Southern blots with three other probes to distinct regions within the open reading frame and obtained similar results (data not shown). These results demonstrate that both Gcnf transcripts are encoded by a single copy gene.
|
Genomic Structure of Gcnf
We have sequenced 80 kb of Gcnf genomic DNA, spanning part of the 5'-UTR, the entire open reading frame, and the complete 3'-UTRs for both Gcnf transcripts (GenBank AF390897–AF390900). The Gcnf coding region contains 11 exons and 10 introns (Fig. 2). Intron-exon boundaries, with typical splice donor and acceptor sites, were determined by comparing genomic DNA and cDNA sequences (Table 1). Nine of the introns range in size from 1.0 to 15.9 kb, and the tenth has a minimum size of 12 kb. The complete DNA sequence was determined for seven of the introns. The exon that encodes the first 33 amino acids of GCNF was designated exon 1. An additional exon upstream of exon 1 (asterisk in Fig. 2) was identified by RT-PCR, although we were unable to confirm this sequence by Northern blot analysis.
|
|
The complete DBD of GCNF is encoded by exon 4 (242 bp, Fig. 2). Unlike most members of the nuclear receptor superfamily, the Gcnf gene has no introns separating the two zinc fingers in the DBD region. Exons surrounding the DBD are quite small, including exons 2 and 3 (42 and 45 bp) and exon 5 (56 bp). Exon 5 encodes the T/A-box domain, a highly conserved 18-amino acid region that mediates specific DNA binding and receptor dimerization [11, 33]. Exons 2–5 are separated by three large introns (>12, 15.9, and 13.5 kb; Fig. 2 and Table 1). The hinge region of GCNF, reported to be essential for interaction with the nuclear corepressor N-CoR [17], is encoded by exon 6 and part of exon 7. Exons 8–11 encode the ligand-binding domain, a region capable of adopting different conformations with distinct dimerization, DNA-binding, and transcriptional properties [33].
Complete Sequence of the 3'-UTRs for Both Gcnf Transcripts
Exon 11 encodes the 3'-UTR regions of both Gcnf mRNAs (Fig. 2 and Table 1). We used 3'-RACE to determine the full length of the 3'-UTR for each transcript. Initially, we obtained two cDNA clones with poly (A) tails at their 3' ends. Both clones hybridized to Gcnf mRNAs on Northern blots. Comparisons of Gcnf cDNA and genomic DNA sequences indicated that one of the clones resulted from annealing of the poly (T) primer to an internal poly (A) sequence in the 3'-UTR of the 7.4 kb transcript (bold type in Fig. 3). Therefore, another primer downstream of this internal sequence was generated to complete 3'-RACE and comparative sequence analysis of the larger transcript.
|
The 3'-UTR of the 7.4 kb transcript is 4451 nucleotides in length, much longer than the average length of 600 nucleotides reported for terminal exons of vertebrate genes [34]. Selected regions of the 3'-UTR sequence are shown in Fig. 3. A conserved polyadenylation signal (AAUAAA) was identified near the end of the 7.4 kb transcript, and polyadenylation appeared to be initiated in two positions. The 3'-UTR sequence of the larger Gcnf transcript was analyzed with UTRscan, a program that searches for UTR functional elements that may regulate mRNA localization, stability, or translational efficiency [35]. This program identified five regions (underlined in Fig. 3) with sequence similarity to tandem repeats referred to as differential control elements. The binding of regulatory proteins to these elements inhibits the translation of erythroid 15-lipoxygenase mRNA [36, 37]. A putative iron-response element was also identified just before the poly (A) tail (double underline in Fig. 3). Iron-response elements located within the 3'-UTR of transferrin receptor mRNA have been implicated in iron-dependent regulation of mRNA stability [38].
The 3'-UTR of the 2.1 kb Gcnf transcript is 202 nucleotides in length, as determined by 3'-RACE (complete sequence shown in Fig. 3). The UTR scan program did not identify any common motifs in this shorter 3'-UTR. Like many other mRNAs expressed in male germ cells, the 2.1 kb transcript does not have a typical polyadenylation signal near the poly (A) tail [39, 40]. Northern blot analysis was used to confirm the 3' end of the 2.1 kb transcript. A probe derived from a common region of the Gcnf sequence (probe 1 in Fig. 4) detected both transcripts, whereas a probe specific for the 7.4 kb mRNA ([4]; probe 3 in Fig. 4) detected only the larger transcript. Probe 2 was designed to hybridize to the region immediately after the predicted 3' end of the 2.1 kb transcript. As predicted by 3'-RACE, this probe detected only the larger 7.4 kb transcript (Fig. 4).
|
Both Gcnf Transcripts Share the Same Coding Region
In our previous Northern analyses, cDNA probes derived from two distinct regions in the open reading frame hybridized to both Gcnf transcripts [4]. These results confirmed that some sequences in the two transcripts are shared, but did not rule out alternative splicing within the coding region. Using exon-specific probes on Northern blots of testicular RNA, we determined that both the 7.4 and 2.1 kb transcripts include sequences encoded by exon 1 and exons 4–11 (data not shown). However, Northern probes specific for exons 2 and 3 were not practical because of the small size of these exons.
To confirm the Northern results and determine if both Gcnf transcripts include sequences from exons 2 and 3, we performed RT-PCR with primers specific for each transcript. The primer sequence specific for the 7.4 kb mRNA (L in Fig. 5A) was derived from the unique 3'-UTR of this transcript (nt 2635–2789 of AF390896). We designed a primer specific for the 2.1 kb transcript based on our 3'-RACE results, since Gcnf sequences restricted to this transcript have not been identified. This primer (GSP-T18; Fig. 5A) anneals to a region including the last eight nucleotides of the 3'-UTR and 18 adenine nucleotides in the poly (A) tail. Reverse transcription was performed with these transcript-specific primers and the cDNAs were amplified by PCR. The resulting RT-PCR products were 2.6 kb in length for the large transcript (Fig. 5B) and 1.6 kb for the small transcript (Fig. 5C), matching the sizes expected from the cDNA without alternative splicing within the coding region. The same RT-PCR conditions used for the GSP-T18 primer were used for RT-PCR with a control primer lacking the poly (T) tail (GSP, Fig. 5A). No products were amplified with the GSP primer (Fig. 5C, lane 3), ruling out the possibility that the 1.6 kb product resulted from reverse transcription of the large Gcnf transcript.
Several independent cDNA clones generated by RT-PCR were sequenced, including 20 clones from each of the transcript-specific reactions and 10 from a reaction with a common primer. Nineteen clones derived from the 7.4 kb mRNA had identical sequences, including all intervening exons. One clone did not contain a sequence from exon 3 (M2 in Fig. 6). Nineteen clones derived from the 2.1 kb transcript also had identical sequences with all exons included. One clone had a short deletion corresponding to the beginning of exon 4 (M3 in Fig. 6). Exons 2 and 3 were deleted in clone M1, derived from RT-PCR with a primer common to both transcripts. However, nine other clones had identical sequences with no deletions. Thus RT-PCR and sequence analysis indicate that Gcnf transcripts in mouse testis share the same open reading frame, but alternative splicing may occur in a small portion of transcripts.
|
Both Gcnf mRNAs Are Translated During Spermatogenesis
We used sucrose gradient sedimentation analysis to determine if both Gcnf transcripts are bound to polysomes in pachytene spermatocytes and round spermatids (Fig. 7). Using a Pgk probe to calibrate the gradients by Northern blot analysis, we detected mRNAs in both the polysomal (fractions 7–10) and nonpolysomal (fractions 1–4) regions of the gradient for each cell type (lower panels, Fig. 7). In the presence of EDTA, these mRNAs dissociate from ribosomes and sediment more slowly than when bound to polysomes. The fractionation of 7.4 kb and 2.1 kb Gcnf mRNAs was monitored with a probe common to both transcripts (upper panels, Fig. 7). Similar to Pgk transcripts, the Gcnf transcripts are localized predominantly in the nonpolysomal fractions in the upper half of each gradient. However, a portion of the 7.4 kb mRNAs sediment in the polysomal regions of the gradient in both pachytene spermatocytes and round spermatids. The 2.1 kb Gcnf mRNA, which is predominantly expressed in round spermatids, is also present in the polysomal fractions. Both 7.4 and 2.1 kb transcripts are dissociated from polysomes in the presence of EDTA and migrate near the top of each gradient. In addition, EDTA reduced the rate of sedimentation of the major peak of Gcnf mRNA from fractions 4 and 5 (gradients containing MgCl2) to fractions 2 and 3.
|
Localization of the GCNF Protein During Spermatogenesis
A GCNF-specific antiserum [15] was used to analyze the expression of this protein in mouse testis. Western blot analysis confirmed the specificity of this antiserum. We detected a single band of the appropriate size (predicted GCNF molecular weight = 55 975) on blots of testicular proteins (Fig. 8, lane 1). Immunoreactivity was blocked completely when the antiserum was preincubated with the peptide immunogen (Fig. 8, lane 2).
|
Sections of adult mouse testis were immunostained with the same antiserum to assess the temporal appearance of GCNF during spermatogenesis and determine its subcellular localization (Fig. 9). The GCNF protein was detected in every tubule cross section (Fig. 9A), primarily in the nuclei of pachytene spermatocytes (p) and round spermatids (r). Less prominent staining in the acrosomes of elongating spermatids (e) and in interstitial tissue between tubules (i) appeared to be nonspecific, since this staining could not be blocked by preincubation of the antiserum with the peptide immunogen (Fig. 9B).
|
As germ cells differentiate, they move from the basal lamina to the lumen of the seminiferous tubule where spermatozoa are released. Because spermatogenesis proceeds with precise kinetics, each cross section of a seminiferous tubule has a distinct cellular association with multiple layers of germ cells developing synchronously. Twelve cellular associations or stages have been identified in the mouse [41]. To examine the relative distribution of GCNF during the meiotic and postmeiotic phases of spermatogenesis, we compared successive stages in the same testis section (Fig. 9, C–J). GCNF immunostaining was not observed in cells closest to the basal lamina, including spermatogonia (g, Fig. 9F); early spermatocytes (c, Fig. 9, G–I); and supporting Sertoli cells (s, Fig. 9, C and F). GCNF was first detected during the pachytene phase of meiosis. Although early pachytene spermatocytes (p) in stage I were unstained (Fig. 9C), GCNF appeared in stage II–IV pachytene spermatocytes within a spherical structure in the nucleus (arrows in Fig. 9, D and E). It is likely that this structure is the XY body, a subregion occupied by the condensed X and Y chromosomes, previously reported to be the only site of GCNF immunostaining in mouse testes [42]. GCNF immunostaining increased during subsequent stages of pachytene spermatocyte differentiation (p in stages V–X, Fig. 9, F–I). During the meiotic divisions in stage XII, condensed chromosomes were intensely labeled with the GCNF antiserum.
The haploid period of differentiation follows the meiotic divisions, beginning with round spermatids in stage I (r, Fig. 9C). GCNF is abundant in round spermatid nuclei (stages I–VIII), particularly in stages I–V (Fig. 9, C–H). Immunostaining is particularly prominent in the heterochromatic region in the center of round spermatids (arrows in Fig. 9, G and H) [43]. As the spermatids begin to elongate in stages IX and X, nuclear GCNF is markedly reduced (e in Fig. 9I).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Using 3'-RACE, we determined the complete sequence of the 3'-UTRs for both Gcnf transcripts. Our results extend the known cDNA sequences for the 2.1 and 7.4 kb transcripts to 1975 and 6137 nucleotides, respectively (see GenBank AF390896). Although the smaller transcript was originally estimated to be 2.3–2.4 kb [1, 2], we have revised the estimated length to 2.1 kb after careful comparisons with markers designed for accurate RNA size determination on Northern blots (Millennium Markers, Ambion). Therefore, we may have obtained the full length cDNA for the small transcript, since several transcripts in round spermatids have poly (A) tracts of 150 bases [46]. However, we were unable to determine the transcription start site of the 2.1 kb mRNA by primer extension assays, despite many attempts. It is likely that these experiments were hampered by the very high GC content (74%) of exon 1, which includes 286 bp of the 5'-UTR. In fact, none of the published mouse and human Gcnf cDNAs include sequences preceding this GC-rich region [1–3, 13, 21, 47, 48]. Consequently, the sequence of the 5' end of the 7.4 kb Gcnf transcript that precedes this region (1.2 kb) remains unknown.
Exons 2 and 3 of Gcnf, which immediately precede the DNA binding domain, are small exons (42 and 45 bp) surrounded by large introns (8.5, >12, and 15.9 kb). Published human GCNF cDNA clones do not contain sequences corresponding to exon 3 in mouse Gcnf [3, 13, 21, 47, 48]. Furthermore, splice variants for human GCNF have been reported with multiple deletions in the N-terminal domain (1; see Fig. 6). To determine if similar alternative splicing occurs in the mouse, we designed primers specific for each of the Gcnf transcripts and analyzed the sequences of RT-PCR-generated clones. These results provide the first evidence that the majority of the 2.1 and 7.4 kb transcripts expressed in mouse testis share the same coding region. Although we did detect small deletions in the region preceding the DNA binding domain (exon 2 through the beginning of exon 4), these deletions occurred in only 6% of the sequenced RT-PCR clones (3/50). Structural elements such as suboptimal splicing signals may account for the observed clustering of Gcnf deletions in this region [49]. For example, truncation of the 5' end of exon 4 (M3 in Fig. 6) may have resulted from the use of a cryptic splice acceptor site, since the deleted fragment (TTTCCGTCCCAG) is very similar to the consensus 3' splice site (polypyrimidine NCAG) [50].
Translational repression is a common feature of gene expression during meiosis and haploid differentiation of spermatids [51], the two phases of spermatogenesis where Gcnf is transcribed [4]. Furthermore, we identified potential regulatory elements in the long 3'-UTR of the 7.4 kb transcript that may contribute to reduced translation or mRNA stability. To determine if both Gcnf transcripts are associated with polysomes, we analyzed cell lysates on sucrose gradients. As reported for the majority of mRNAs in pachytene spermatocytes and round spermatids [51], significant fractions of the 2.1 and 7.4 kb transcripts were found in nonpolysomal fractions of both cell types. However, a portion of both Gcnf transcripts were associated with polysomes, suggesting that both are translated during spermatogenesis. In addition, the observed effect of EDTA on the slowly sedimenting major peak of Gcnf transcripts suggests that unusual features of these mRNAs such as upstream reading frames may promote association with monosomes and small polysomes [46].
A previous study using a different antibody detected GCNF protein in isolated mouse pachytene spermatocytes and round spermatids on Western blots, but only in the XY body of pachytene spermatocyte nuclei in immunofluorescence assays [42]. In our studies GCNF was distributed throughout the nucleus of both pachytene spermatocytes and round spermatids. GCNF was first detected in stage II–IV pachytene spermatocytes in a nuclear structure consistent with the XY body, which contains the condensed X and Y chromosomes. We also observed high levels of GCNF in condensed chromosomes undergoing the meiotic divisions and in the central heterochromatic region in round spermatid nuclei. The prominence of GCNF in heterochromatic regions and meiotic chromosomes is consistent with reports that this orphan receptor functions as a transcriptional repressor [6, 16, 17, 33].
Protamine 1 and protamine 2, which have DR0 response elements in their promoters, have been identified as potential targets for GCNF regulation [15, 19]. The mRNAs for these sperm nuclear proteins are transcribed in round spermatids [52], but are not translated until step 11–12 spermatids [53, 54]. Since GCNF levels are high when protamine mRNAs begin to accumulate in round spermatids, it seems unlikely that GCNF inhibits the transcription of the protamines during these stages. However, interactions of GCNF with other proteins or cross-talk between nuclear receptor signaling pathways [17] could modulate the expression of the protamine genes during spermatogenesis.
In summary, this study provides additional information on potential regulatory regions of Gcnf including quite diverse 3'-UTR sequences for the two transcripts, confirms that the same protein is translated from both transcripts, and defines the stages of spermatogenesis where the GCNF protein is expressed. These results suggest that GCNF plays a transcriptional regulatory role during spermatogenesis, from the pachytene stage of meiotic prophase through the early phase of spermatid differentiation. Additional studies are needed to identify other genes that are potential GCNF targets during these stages and to define mechanisms that regulate the transcription and translation of the two Gcnf mRNAs during spermatogenesis.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Deborah A. O'Brien, Department of Cell and Developmental Biology, CB# 7090, 214A Taylor Hall, University of North Carolina School of Medicine, Chapel Hill, NC 27599-7090. FAX: 919 966 1856; dao{at}med.unc.edu
3 Current address: Vanderbilt-Ingram Cancer Center, 512 Preston Research Building, Vanderbilt University, Nashville, TN 37232-6838
Received: 9 October 2002.
First decision: 31 October 2002.
Accepted: 21 November 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) in rat androgen-dependent tissues and human prostate cancer. J Androl 2001 22:471-480[Abstract]This article has been cited by other articles:
|
|
 |
|
  Z.-J. Lan, X. Xu, A. C.-K. Chung, and A. J. Cooney Extra-Germ Cell Expression of Mouse Nuclear Receptor Subfamily 6, Group A, Member 1 (NR6A1) Biol Reprod, May 1, 2009; 80(5): 905 - 912. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. L. Sartini, H. Wang, W. Wang, C. F. Millette, and D. L. Kilpatrick Pre-Messenger RNA Cleavage Factor I (CFIm): Potential Role in Alternative Polyadenylation During Spermatogenesis Biol Reprod, March 1, 2008; 78(3): 472 - 482. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Yan, Y. Lu, H. Sun, D. Tao, S. Zhang, W. Liu, and Y. Ma A microarray for microRNA profiling in mouse testis tissues Reproduction, July 1, 2007; 134(1): 73 - 79. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Mikawa, T. Morozumi, S.-I. Shimanuki, T. Hayashi, H. Uenishi, M. Domukai, N. Okumura, and T. Awata Fine mapping of a swine quantitative trait locus for number of vertebrae and analysis of an orphan nuclear receptor, germ cell nuclear factor (NR6A1) Genome Res., May 1, 2007; 17(5): 586 - 593. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Benoit, A. Cooney, V. Giguere, H. Ingraham, M. Lazar, G. Muscat, T. Perlmann, J.-P. Renaud, J. Schwabe, F. Sladek, et al. International Union of Pharmacology. LXVI. Orphan Nuclear Receptors Pharmacol. Rev., December 1, 2006; 58(4): 798 - 836. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Wang, B. L. Sartini, C. F. Millette, and D. L. Kilpatrick A Developmental Switch in Transcription Factor Isoforms During Spermatogenesis Controlled by Alternative Messenger RNA 3'-End Formation Biol Reprod, September 1, 2006; 75(3): 318 - 323. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Rajkovic, R. Middendorff, M. G. Wetzel, D. Frkovic, S. Damerow, H. J. Seitz, and J. M. Weitzel Germ Cell Nuclear Factor Relieves cAMP-response Element Modulator {tau}-mediated Activation of the Testis-specific Promoter of Human Mitochondrial Glycerol-3-phosphate Dehydrogenase J. Biol. Chem., December 10, 2004; 279(50): 52493 - 52499. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Nakajima, S. Fujino, G. Nakanishi, Y.-S. Kim, and A. M. Jetten TIP27: a novel repressor of the nuclear orphan receptor TAK1/TR4 Nucleic Acids Res., August 9, 2004; 32(14): 4194 - 4204. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
