| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
a Department of Cell Biology, University of Virginia, Charlottesville, Virginia 22908
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The 5' flanking sequences for a limited number of testis-specific genes have been cloned, and the minimal promoters required for testis-specific gene expression have been defined in transgenic mice. These include phosphoglycerate kinase (Pgk-2), proenkephalin, acrosin, histone H1t, pyruvate dehydrogenase (Pdha-2), and lactate dehydrogenase (Ldhc-4) genes [1–6], which are expressed in spermatocytes during the meiotic phase, and protamine 1, protamine 2, and testis angiotensin-converting enzyme (t-ACE) genes [7–9], which are expressed in the late spermatids. Promoter analysis of genes uniquely expressed in early round spermatids has not been previously reported, to our knowledge.
Recent studies involving generation of knockout mice demonstrated critical roles for the testis-specific transcription factors A-myb and CREM
in coordinate gene expression during the meiotic phase and spermiogenesis, respectively [10–12]. However, early spermatid gene expression was not affected in the CREM knockout mice [11], suggesting a role for a CREM-independent transcriptional control mechanism in the initial part of spermiogenesis.
SP-10 is an intra-acrosomal protein first identified in human sperm [13] and subsequently shown to be present in the acrosomes of monkey, baboon [14], fox [15], mouse [16], and bull [17] sperm. Following the acrosome reaction, some amount of the SP-10 protein is retained on the inner acrosomal membrane [18]. The finding that antibodies to SP-10 block sperm-egg interactions in vitro [19] suggested a role for SP-10 during fertilization.
In the human seminiferous epithelium, SP-10 mRNA was first detected in early Golgi phase step 1 spermatids during formation of the acrosomal granule [20]. An extensive tissue specificity analysis indicated that SP-10 transcription is testis-specific [21]. The SP-10 gene thus provides a useful model for study of the mechanism of testis-specific gene transcription during acrosome biogenesis in round spermatids. Here we report on the round spermatid-specific expression, genomic cloning, and organization of the mouse SP-10 (mSP-10) gene, as well as delineation of its functional promoter in transgenic mice. The study demonstrates that the proximal promoter of the mSP-10 gene, including 266 base pair (bp) of genomic sequence upstream of the germ cell-specific transcriptional start site and 28 bp of the flanking downstream sequence, contains the promoter elements necessary to mediate round spermatid-specific gene expression.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The full-length mSP-10 cDNA [16] was used as a probe to screen a mouse liver (C57BL/6N male) genomic library (ML1043j) constructed in EMBL phage vector (Clontech, Palo Alto, CA). Plaque hybridization was performed as previously described [22]. Briefly, the library was plated at a density of 105 plaque forming units (pfu)/150-mm plate using Escherichia coli K802 (Clontech) as the bacterial host strain, and duplicate plaque lifts were transferred to Duralon-UV nylon filter circles (Stratagene, La Jolla, CA). Prehybridization of the filters was carried out for 2 h at 42°C in a buffer containing 50% formamide (Sigma Chemical Co., St. Louis, MO), 6-strength SSC (single-strength: 0.015 M sodium citrate, 0.15 M sodium chloride), 5-strength Denhardt's solution (single-strength: 0.5% each of Ficoll, BSA, and polyvinyl pyrrolidone), and 1% SDS. The mSP-10 cDNA was labeled with [32P]dCTP using the Prime-a-Gene kit (Promega, Madison, WI). Hybridization was performed at 42°C for 20 h in the prehybridization buffer consisting of 106 cpm/ml of radiolabeled mSP-10 cDNA probe. The filters were washed in a buffer containing double-strength SSC and 1% SDS at room temperature for 30 min, at 65°C for 1 h, and at 65°C in 0.2-strength SSC and 1% SDS for 1 h. The dried filters were exposed to autoradiographic film (Eastman Kodak, Rochester, NY) at -70°C. Only signals obtained in duplicate were considered positive, and the corresponding plaques were cored out from the plates. For plaque purification, secondary and tertiary screenings were done at a density of 102-103 pfu/150-mm plate.
Mouse SP-10 Genomic Clones
A total of 20 genomic phage clones that hybridized with the mSP-10 cDNA were plaque purified. Clone 
3, containing the largest insert (14 kilobases [kb]), was further analyzed. Oligonucleotide probes originating from the 5' untranslated region (UTR; 5'tcttctcagctcttgagtgtgccac, position 11–35, GenBank accession #U31992) and 3' UTR (5'gctctgacttaggcaggttcaccac, position 871–895, GenBank accession #U31992) of mSP-10 cDNA were [
32P]ATP end-labeled using T4 kinase (Promega) according to the manufacturer's instructions. The two radiolabeled oligonucleotide probes were used separately to hybridize to the clone 3 DNA (Sambrook et al. [22]). Both probes hybridized with the clone 3 DNA, indicating that the 14-kb insert contained the entire mSP-10 genomic sequence. Restriction mapping revealed three internal EcoRI sites, splitting the 14-kb insert of clone 3 into 4.7-, 3.8-, 3.2-, and 2.2-kb fragments that were subcloned separately into Bluescript vector to yield p1.4, p2.1, p3.4, and p4.1, respectively. The orientation of the genomic subfragments was determined based on the hybridization of oligonucleotides derived from mSP-10 cDNA as well as DNA sequencing (see below). The 5' UTR oligonucleotide probe (position 11–35, GenBank accession #U31992) hybridized to p4.1, suggesting that the 2.2-kb EcoRI fragment contained the 5' end of the gene. Similar analysis using the 3' UTR sequence of SP-10 cDNA (position 871–895, GenBank accession #U31992) placed the p1.4 insert at the 3' end of the gene. Several probes chosen from the middle portion of the cDNA hybridized to the p4.1 clone. Clones p2.1 and p3.4 did not hybridize with probes derived from mSP-10 cDNA.
DNA Sequencing
Nucleotide base sequences of the bluescript plasmids p4.1, p1.4, p3.4, and p2.1 containing the SP-10 genomic DNA were obtained by Sanger's dideoxy chain termination method [22]. Initially, the vector-based T3 and T7 primers were used, and subsequently new internal primers were synthesized to facilitate further sequencing. In particular, the regions corresponding to exon-intron junctions and the 5' and 3' UTR were sequenced; however, entire intronic regions were not sequenced.
Primer Extension Analysis
A 29-mer primer sequence (5'acccagtaagattaactccttcattttga) complementary to the 60- to 88-bp region of SP-10 cDNA (GenBank accession #U31992), which included the ATG site, was used to reverse transcribe mRNAs from mouse testis and liver. Testis and liver mRNAs (500 ng) were annealed to 100 fmol of [32P]ATP end-labeled 29-mer primer. First-strand cDNA synthesis was performed using the AMV Reverse Transcriptase Primer Extension System (Promega) according to the manufacturer's instructions. Primer extension products were analyzed by polyacrylamide gels containing 8% acrylamide, 7 M urea in single-strength TBE (89 mM Tris, 110 mM boric acid, 2 mM EDTA, pH 8.0). Standard M13 dideoxy chain termination sequencing reaction products were run alongside to provide molecular size markers. The length of the primer extension products reflected the total number of bases between the labeled nucleotide of the primer and the 5' end of the RNA.
Reporter Gene Constructs
Three promoter fragments of mSP-10, whose 5' ends started from -408, -266, or -91 bp and whose 3' position terminated at bp +28 with reference to the transcription start site, were cloned upstream of green fluorescent protein (GFP) cDNA using reporter vector pEGFP1 (Clontech). The -408 to +28-bp and -91 to +28-bp mSP-10 fragments were amplified by polymerase chain reaction (PCR) from mouse genomic lambda clone 3 using as 5' end primers 5'cccctcgagcctccaatcttaggactaacctcag and 5'cccctcgagaagaggaacaacccattgtga (XhoI site in italics), respectively, and a common 3' end primer with the sequence 5'gggggatcctggcacactcaagagctgaga (BamHI site in italics). Vent DNA polymerase (New England Biolabs Inc., Beverly, MA) was used in 25 cycle reactions. The PCR products were cleaved with XhoI and BamHI and ligated to XhoI+BamHI-cleaved pEGFP1 vector to yield -408SP10-gfp and -91SP10-gfp constructs that were verified by DNA sequencing. The -408SP10-gfp and -91SP10-gfp plasmids were cleaved with XhoI and AflII (pEGFP1, bp#1058) to generate the -408SP10-gfp and -91SP10-gfp transgenes, each consisting of an mSP-10 promoter segment, GFP, and SV40 polyadenylation signals. The third -266SP10-gfp transgene was derived by cleavage of the -408SP10-gfp clone using HincII (at -266 in mSP-10, see see Fig. 5 and AflII, which released a 1352-bp fragment. The -408, -266, and -91 transgenes were gel purified prior to injection into mouse eggs.
|
Generation of Transgenic Mice
Transgenic mice were produced by microinjection of transgenes into the male pronuclei of fertilized mouse eggs and transfer of these eggs to pseudopregnant foster mothers using standard procedures [23] in the Transgenic Mouse Core Facility at the University of Virginia. Transgenic founder mice were identified by two methods. 1) Tail DNA was subjected to PCR amplification using GFP primers, 5'gtgagcaagggcgaggagctg (nucleotides [nt] #100–120, GenBank accession #U55761) and 5'cttgtacagctcgtccatgccgag (nt #813–790, GenBank accession #U55761) spanning a 713-bp portion of the GFP cDNA. Mouse SP-10 genomic DNA primers (5'cctccaatcttaggactaacctcag and 5'tggcacactcaagagctgaga) were used as a positive internal control. The amplification of 713-bp gfp PCR product identified the founder mice. 2) Transgene integration was confirmed by Southern hybridization [22]. Mouse genomic DNAs were cut with EcoRI, transferred to nylon membrane (Stratagene), and probed with radiolabeled GFP cDNA (derived from pEGFP1). The filters were washed under the same conditions as described for screening of the mouse genomic library. The founder mice were identified on the basis of positive hybridization band(s) obtained with the gfp probe. The positive founders were mated with C57 Bl partners to obtain an F1 generation for analysis. Adequate numbers of founder transgenic mice were generated for each construct to allow interpretation of the outcome.
GFP Expression in Transgenic Mice
F1 transgenic males (6 to 8 wk old) were killed according to the animal use guidelines provided by the institution, and tissues were collected in sterile PBS to study the expression of GFP. The testes of transgenic and nontransgenic littermates were decapsulated and observed using a Zeiss (Oberkochen, Germany) Axiovert fluorescence microscope with fluorescein isothiocyanate (FITC) filters [24]. To prepare cell squashes, 2- to 3-mm pieces of seminiferous tubules were cut and transferred to a glass slide, and a coverslip was then placed on the preparation to obtain a monolayer of cells. The seminiferous tubules expressing GFP appeared bright green when viewed using the FITC optics. Somatic tissues including brain, liver, lung, heart, kidney, muscle, intestine, spleen, seminal vesicle, and prostate from transgenic and nontransgenic mice were also processed and examined for GFP production.
Reverse Transcription-Polymerase Chain Reaction (RT-PCR)
Poly(A)+ RNAs isolated from testes and several somatic tissues of transgenic animals were reverse transcribed using an RT-PCR kit (Stratagene) according to the manufacturer's instructions. The poly(A)+ RNA samples were routinely treated with DNase and extracted with phenol/chloroform prior to use in RT-PCR. PCR was then performed using the GFP-specific primers described above. Production of a 713-bp PCR product indicated the presence of transgene mRNA. PCR using mouse beta-actin primers (5'gtgggccgctctaggcaccaa, and 5'ctctttgatgtcacgcacgatttc) served as an internal control. Appearance of a 500-bp beta-actin PCR product indicated that the reaction conditions for RT and PCR were appropriate.
In Situ Hybridization
The testes of sexually mature mice were collected in 10% neutral buffered formalin (Sigma). Processing of tissue samples and sectioning were carried out as previously described [25]. In situ hybridization was performed using mSP-10 and GFP riboprobes. For the purpose of in vitro synthesis of riboprobes, mSP-10 and GFP cDNAs were cloned into the bluescript vector (Stratagene). The mSP-10 cDNA was excised from pETmSP-10 [16] using EcoRI and BamHI and was cloned into the vector to yield pBS-SP10. The pEGFP1 vector was cut with KpnI and NotI to release GFP cDNA, which was then ligated to KpnI+NotI-treated pBSII KS- vector to yield pBS-gfp. Radiolabeled sense and antisense mSP-10 and GFP probes were generated from linearized plasmids by incorporating [5–3H]CTP and [5,6–3H]UTP (Amersham, Piscataway, NJ) using the Riboprobe Combination System (Promega). The hybridization and wash conditions were as described previously [25].
Detection of GFP in the Developing Testis
Male pups born to the -266SP10-gfp mice were killed at daily intervals between Days 14 and 35 postpartum. The day of birth was considered Day 1. Testes were collected in sterile PBS; the seminiferous tubules were teased, and 2- to 3-mm sections were cut. The spermatogenic cell types were identified according to established criteria [26] under phase contrast. Photographs were obtained using a digital camera (Hamamatsu, Bridgewater, NJ) and Adobe Photoshop (San Jose, CA) software.
Database Searches
The mouse SP-10 5' flanking sequence was subjected to a database search to identify putative transcription factor binding sites, using the Tfsites program (Genetics Computer Group, Madison, WI).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In situ hybridization was performed to identify cell types expressing mSP-10 mRNA during spermatogenesis. In the mouse, the cycle of the seminiferous epithelium has been divided into 12 consecutive stages, each stage characterized by a constant cell association [26]. In a typical cross section of mouse testis, seminiferous tubules representing various stages of the cycle are observed (Fig. 1). The mSP-10 mRNA hybridization signal was restricted to seminiferous tubules at stages I through VII of the cycle; there was no mSP-10 expression at stages VIII–XII. An example of mSP-10 expression in the mouse germinal epithelium is shown in Figure 1. The mSP-10 mRNA hybridization signal was abundant at stage III, which is characterized by the presence of numerous round spermatids and a single layer of pachytene spermatocytes, but it was greatly diminished at stage VII of the cycle (Fig. 1A), in which elongate spermatids (s16) typically line the luminal surface of the seminiferous epithelium (Fig. 1B). Mouse SP-10 mRNA expression was seen predominantly over the cytoplasm of the round spermatids at stage III, whereas in all other germ cell types including spermatogonia, spermatocytes, or late spermatids at either stage (Fig. 1A), the hybridization signal was not above the background level obtained by probing with the mSP-10 sense riboprobe (data not shown). The in situ data indicate that mSP-10 mRNA expression is transient and stage-dependent and is highly expressed in early round spermatids, coincident with acrosome biogenesis. Expression of the mSP-10 gene thus represents an example of testis-specific transcription activity during the initial phase of spermiogenesis.
|
Genomic Organization of mSP-10 Gene
Subcloning and sequencing of a 14-kb genomic DNA insert showed that the mSP-10 gene consisted of four exons spread over a 5.5-kb region (Fig. 2). Exons 1 through 4 corresponded to nt 1 through 107, 108 through 599, 600 through 719, and 720 through 1067, respectively. Nucleotide sequences of these exons were identical to the mouse SP-10 cDNA (GenBank accession #U31992). Introns 1 through 3 were determined to be 0.8, 1.9, and 1.7 kb in length (Fig. 3). The sequences at the exon-intron boundaries of the mSP-10 gene were consistent with the higher eukaryotic splice sequences that aid in the removal of introns. The 5' splice donor and 3' splice acceptor sequences of mSP-10 gene conformed to the consensus eukaryotic splice site sequences AG\gtaag (5' donor) and cag/G (3' acceptor). Similarly, a canonical branch point sequence (consensus YNCURAY) and a poly pyrimidine tract [27] were located at appropriate positions upstream of the 3' splice acceptor site of each intron of the mSP-10 gene (Fig. 3). It is interesting to note that the introns consistently interrupted the exons of the mSP-10 gene between the first and second bases of a codon.
|
|
Figure 2 depicts the exon organization in relation to the structure of the SP-10 protein. Each exon encoded a structural part of the SP-10 protein previously characterized in the human and mouse [28, 29]. Exon 1 included 59 bp of 5' UTR and the first 17 amino acids of the SP-10 protein, which constitute a hydrophobic signal peptide. Exon 2 encoded the 164 amino acid hydrophilic core of the protein. Exons 3 and 4 included the carboxyl terminus of the protein, which has been shown to be conserved among the mouse, human, baboon, and macaque SP-10 cDNAs [14].
Transcriptional Start Site of mSP-10
The transcriptional start site of the mSP-10 gene was determined by the primer extension method. A radiolabeled 29-mer oligonucleotide primer complementary to nt -5 to +24 bp relative to the ATG codon (see Fig. 5, broken underline) was hybridized to mRNAs from mouse testis and liver, and the first-strand synthesis was carried out in the presence of avian myeloblastosis virus reverse transcriptase. The reaction products were analyzed on a denaturing sequencing gel alongside a size marker (Fig. 4). One major extension product (arrow) of 82 bases was obtained with testis (T) but not with liver (L) mRNA (Fig. 4). On the basis of the length of the primer extension product, the adenosine residue located 58 nt upstream of the ATG was determined to be the transcription initiation site (designated +1, Fig. 5). None of the mSP-10 cDNA clones sequenced to date contained additional sequence upstream of the +1 site as determined by this method, indicating that mSP-10 transcription is initiated from a single site.
|
5' Flanking Sequence of SP-10 Gene
The core promoter of the mouse SP-10 gene lacked a canonical TATA box in the -25-bp position but contained a pyrimidine-rich initiator (Inr) sequence in the -2 to +5-bp position, encompassing the transcription start site (Fig. 5). The mSP-10 Inr sequence (5'TCA+1GT+3TT) conformed to the consensus eukaryotic Inr sequence 5'YYANTYY (Y is a pyrimidine; N is any nucleotide), in which the +1 and +3 positions are invariably occupied by A and T, respectively [30]. In eukaryotic promoters that lack a TATA box, the Inr element has been shown to be sufficient for accurate transcription initiation [30]. The human SP-10 promoter also lacks a TATA box and, like the mouse gene, contains a consensus Inr element surrounding the transcriptional start point ([31], Fig. 5). A comparison of the mouse SP-10 promoter sequence with the previously published 5' flanking sequence of the human SP-10 gene ([31]; GenBank accession #S65606s1) revealed that the -408 to +58-bp region of the mouse gene shared 67% identity and 80% similarity with the -459 to +67-bp human gene sequence (Fig. 5). In contrast, regions further upstream in both gene promoters shared less than 50% homology (data not shown).
Several identical cis sequences were shared by the two promoters (highlighted regions in Fig. 5); some of these constituted recognition sites for known transcription factors. A CAT box, typically located close to the transcription start site in eukaryotic promoters, was found at -404 bp in the mouse and at -455 bp in the human SP-10 promoter. A consensus GATA-1 binding sequence was located at -242 in the mouse (5'CTATCT) and at -280 (5'AGATAT) in the human. The transcription factor GATA-1, which plays a critical role in the differentiation of the hematopoietic system, has also been shown to be highly expressed in the testis [32]. A consensus Ets-1 binding site, 5'CAGGATGT, was located at -99 bp in the mouse. The human SP-10 promoter also contained an Ets-1 site, 5'CAGG
T
C at -110 bp, but this deviated from the consensus at the two positions underlined. A 5'-CCCC sequence, which has a high affinity for the ETF transcription factor, is conserved at -169 and -200 in the mouse and human, respectively. The ETF factor preferentially activates transcription from TATA-less promoters [33].
One notable feature of the proximal SP-10 promoter region is the conservation of three palindrome sequences (Fig. 5). Palindromes P1, P2, and P3 positioned at -47 (5'TTCTAGAA), -105 (5'CCTGaaCAGG), and -141 bp (5'CACTAGTG) in the mouse, were conserved at -49, -116, and -157-bp positions, respectively, in the human (shaded boxes in Fig. 5). It is well known that palindrome sequences constitute the half sites recognized by transcription factors, but no recognition sites for known transcription factors were found within the SP-10 palindromes. A 16-bp T-rich region was conserved at -202 and -231 positions in the mouse and human, respectively. There were no GC-rich elements or Sp1 binding sites in the SP-10 promoter.
The mouse and human SP-10 promoters contain a sequence similar to the consensus cAMP response element (CRE, 5'TGACGTCA) at -74 (5'TGAGGACA) and -76 (5'TGAAGAAA) positions, respectively. The cognate transcription factor CREM (cAMP response element modulator tau) was previously shown to be essential for the transcription of a number of testis-specific genes that contain the CRE sequence in their proximal promoter. However, CREM-deficient mice continued to produce SP-10 mRNA in the testis (Sassoni Corsi, personal communication), indicating that CREM is probably not involved in SP-10 transcription. In addition to the shared promoter elements, there were some transcription factor binding sites unique to the mouse promoter, including an E-box site (5'CATGTG) at -221, HNF-5 (5'TGTTTGT) at -198, and AP1 site (5'TGATTCA) at -27 positions (Fig. 5). The presence of an E-box is particularly noteworthy in view of the recently identified germ cell-specific, E-box binding transcription factor, FIG (factor in the germline)-alpha [34].
Generation of Transgenic Mice
The high degree of similarity between mouse and human sequences observed in the SP-10 proximal promoter region (Fig. 5) may indicate conservation of functionally important cis-acting elements responsible for testis-specific transcription. The putative regulatory elements within the conserved -408 to +28-bp mSP-10 promoter region are schematically represented in Figure 6A. To identify the minimal mSP-10 promoter region required to drive testis-specific gene expression in transgenic mice, three separate reporter gene constructs were made. An mSP-10 promoter containing the 1) -408 to +28-bp, 2) -266 to +28-bp, or 3) -91 to +28-bp fragment was linked to GFP cDNA (Fig. 6B). Transgenic founder mice were identified by PCR amplification of the GFP sequence, and the integration of the transgene in each line was confirmed by Southern hybridization (see Materials and Methods). Three founders containing the -408 to +28-bp fragment, six founders containing the -266 to +28-bp fragment, and three founders containing the -91 to +28-bp fragment were identified.
|
Testis-Specific Expression of GFP in Transgenic Mice
F1 transgenic males, 6 to 8 wk old, were killed, and tissues were examined for reporter gene expression. Since the fluorescence emanating from GFP-expressing cells could be visualized using a microscope fitted with the FITC filter set, freshly isolated tissue samples were directly examined to detect reporter gene activity. Tissues obtained from nontransgenic littermates served as negative controls. Fluorescence microscopy demonstrated GFP expression in the seminiferous tubules from mice containing the -408SP10-gfp and -266SP10-gfp constructs, but not the -91SP10-gfp construct (Fig. 6B). Cell squashes prepared from multiple somatic tissues including brain, liver, lung, heart, kidney, muscle, intestine, spleen, seminal vesicle, and prostate of all three transgenic lines were negative for GFP expression (data not shown), indicating that the transgene expression in -408SP10-gfp and -266SP10-gfp mice was testis specific. The -91SP10-gfp mice failed to express GFP in any tissue.
The testis specificity of GFP expression was confirmed by RT-PCR. Messenger RNA obtained from various tissues of transgenic mice was reverse transcribed and subjected to PCR amplification using GFP- and beta-actin-specific oligonucleotide primers (see Materials and Methods). The 713-bp PCR product indicative of the presence of GFP mRNA was detected in the testes of the -408SP10-gfp and -266SP10-gfp mice, but not the -91SP10-gfp mice (Fig. 7). No GFP RT-PCR product was seen in RNA samples representing brain, liver, kidney, heart, or prostate tissues obtained from any of the three transgenic lines. On the other hand, the 500-bp product corresponding to beta-actin control was amplified by all the RNA samples examined (Fig. 7). These results indicated that the -408 to +28-bp or -266 to +28-bp mSP-10 promoter fragments contained the cis-elements required to drive testis-specific expression of a reporter gene in transgenic mice. In contrast, the -91 to +28-bp mSP-10 region by itself was insufficient to activate gene expression in any tissue. It is interesting to note that recognition motifs for GATA-1, bHLH proteins, HNF-5, and Ets-1, as well as two conserved palindrome sequences P3 and P2, were located within the -266 to -92-bp region of the mSP-10 promoter (Fig. 6A).
|
Of three -408SP10-gfp founder mice, two expressed GFP in a testis-specific manner, but the third showed no GFP expression in any tissue, possibly owing to a position effect. Five of six -266SP10-gfp founder lines expressed GFP in the testis alone, whereas all three -91SP10-gfp lines revealed no trace of transgene expression in any tissue.
Spatial and Temporal Pattern of Transgene Expression
In situ hybridization was performed on serial cross sections of the testis of a -408SP10-gfp transgenic mouse using SP-10 and GFP riboprobes to compare the spatiotemporal distribution of transgene mRNA in the seminiferous epithelium with that of the endogenous SP-10. As expected, the SP-10 antisense probe hybridized with only a subset of tubules in the cross section (Fig. 8A), reflecting the stage specificity observed previously. More important, the GFP antisense probe produced a very similar hybridization pattern in an adjacent cross section (Fig. 8B). Thus, the GFP mRNA expression was confined to the same stages of the seminiferous epithelium at which the endogenous mSP-10 gene was also expressed. Similarly, the lack of transgene expression was coincident with the absence of SP-10 expression (compare panels A and B in Fig. 8). These results indicate that the -408 to +28-bp mSP-10 promoter fragment contains the necessary transcription regulation signals responsible for both temporal and testis-specific gene expression in transgenic mice. In general, the in situ hybridization signal corresponding to GFP mRNA was less intense than that of SP-10 (compare panels A and B in Fig. 8). One interpretation of these data is that cis-acting elements located outside of the -408 to +28-bp SP-10 genomic region may be necessary to attain promoter strength/efficiency comparable to that of the endogenous gene. Alternatively, the GFP mRNA may be less stable in the mouse testis.
|
-266 to +28-bp mSP-10 Promoter Was Sufficient to Direct Round Spermatid-Specific Gene Expression In Vivo
We sought to determine precisely the germ cell type in which the transgene expression was initiated by searching for GFP in the developing testis of the transgenic mice, taking advantage of the fact that germ cells of the mouse undergo synchronous differentiation during the first spermatogenic cycle in the developing testis. It is known that type A and type B spermatogonia appear by Day 8, early and late pachytene spermatocytes by Days 14 and 18, respectively, and haploid spermatids between Days 18 and 20 postpartum in the developing testis [35]. Therefore we examined the seminiferous tubules of -266SP10-gfp transgenic mice ranging from 14 to 35 days in age at daily intervals in order to detect the first appearance of GFP fluorescence. Nontransgenic littermates served as negative controls (data not shown). The seminiferous tubules of 18- and 21-day-old transgenic mice are shown in Figure 9 to illustrate the transgene expression pattern. No GFP expression was seen in the pachytene spermatocytes of the Day 18 testis (Fig. 9, Ps, GFP). GFP was first detected over round spermatids, which made their appearance in the Day 21 testis (Fig. 9, Rs, GFP), indicating that the transgene expression was initiated postmeiotically. Spermatocytes at Day 21 remained negative for GFP expression, indicating that the SP-10 promoter is quiescent in germ cells undergoing meiosis (Fig. 9, Ps, GFP). The data from this developmental series clearly indicate that the -266 to +28-bp mSP-10 promoter region activated gene expression during the early haploid phase of spermatogenesis in vivo.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The mouse SP-10 gene, like its human counterpart [20], exhibited a germ cell type- and stage-dependent pattern of gene expression. Transcription of the mSP-10 gene was abundant in postmeiotic round spermatids at stage III of the cycle of seminiferous epithelium, and the mRNA signal rapidly declined in spermatids at stage VII. The SP-10 gene therefore is a good model for the study of round spermatid-specific transcriptional regulation.
The mSP-10 gene promoter lacks a consensus TATA box in the -25 region. Primer extension identified one major transcription initiation site. The sequence surrounding this transcriptional start point, TCAGTTT (where A is the +1), was consistent with that of the consensus Inr sequence first identified in the promoter of the gene for terminal deoxy transferase (TdT) by Smale and Baltimore [30]. Unlike other testis-specific genes lacking TATA boxes, such as acrosin and Ldhc-4 that initiate from multiple sites [3, 36], the SP-10 gene utilized only one major transcription start site embedded within the Inr sequence. This indicates that core promoter elements other than TATA may play a role in accurate initiation of SP-10 transcription. In higher eukaryotes, Inr was shown to be a core promoter element capable of recruiting the preinitiation complex and determining the transcription start site in the absence of a TATA box [37]. It is interesting to note that the transcription start site of human, baboon, and macaque SP-10 genes also mapped within a consensus Inr sequence [14, 31]. The conserved SP-10 Inr, which is similar to the Inr elements of TdT or adenovirus major late promoter gene [30], may serve to assemble the basal transcription complex, while further-upstream promoter elements may be responsible for mediating testis-specific transcription of SP-10 mRNA.
On the basis of the observation that a 408-bp sequence upstream of the transcription start site of the mouse SP-10 gene shared 80% similarity with a corresponding region in the human SP-10 promoter, we predicted that functionally important promoter elements might be conserved within this region. Transgenic mice demonstrated that the -408 to +28-bp mSP-10 promoter fragment contains the necessary promoter elements to drive testis-specific expression of a reporter gene. Furthermore, the promoter activity of the -408 to +28-bp fragment mimicked that of the endogenous mSP-10 gene, in that the spatiotemporal pattern of transgene transcription was comparable to that of SP-10.
Transgenic mice bearing a truncated -266 to +28-bp mSP-10 fragment also expressed GFP mRNA in the testis alone, thus delineating the critical part of the promoter to a smaller region. The -266 to +28-bp promoter retained the ability to direct developmental stage-specific expression of the reporter gene. The appearance of the GFP protein in the 21-day-old, but not the 18-day-old, testes of the transgenic mice is consistent with promoter activity of the -266 to +28-bp fragment in postmeiotic round spermatids. When the SP-10 promoter was truncated further from the 5' end, however, the SP-10 promoter lost its activity, as mice bearing the -91 to +28-bp SP-10 core promoter failed to express the transgene in any tissue. One interpretation of these data is that the general transcription machinery recruited by the SP-10 core promoter must interact with transcription factors recognized by the -266 to -91-bp region in order to activate SP-10 gene transcription.
Taken together, the experiments on transgenic mice indicated that the cis-acting elements, which play a critical role in early round spermatid gene expression, reside within the -266 to +28-bp region of the SP-10 promoter. It is interesting to note that sites for GATA-1, E-box binding protein (-220), HNF-5 (-196 bp), and Ets-1 (-99 bp) are located within the -266 to -91-bp region of the SP-10 promoter, as are the two conserved palindrome sequences, P2 and P3 at -105 and -141 positions, respectively. When compared to the promoters of other testis-specific genes expressed in the round spermatids, including acrosin [3], the -408 to +28-bp SP-10 promoter did not show any significant homology. The cis-acting elements shown to be indispensable for the testis-specific expression of reporter genes in vivo, such as the CTCCAG repeats within the GCP1 sequence of proenkephalin gene [38], the 31-bp palindrome of the lactate dehydrogenase c [6], or the TE element of histone H1t gene promoter [39], were not shared by the mSP-10 promoter identified in this study. It should be borne in mind, however, that the transcription of the mouse acrosin, proenkephalin, Ldhc-4, and H1t genes is initiated during the meiotic phase in spermatocytes whereas SP-10 gene expression is initiated postmeiotically in round spermatids. Therefore the cis-regulatory elements involved in SP-10 gene transcription may be expected to differ.
Recent studies identified A-myb and CREM as transcription factors critical for spermatocyte and spermatid differentiation, respectively. While there is no A-myb recognition sequence in the mSP-10 promoter, one CRE-like element is located at the -74 position. However, we have shown in this study that transgenic mice bearing the -91 to +28-bp promoter, which included the CRE-like element, failed to express GFP in any tissue. Furthermore, the CREM knockout mice continued to express SP-10 mRNA in the testes (Sassoni Corsi, personal communications). This is in accord with the observation that the formation of step 1 spermatids, in which SP-10 transcription begins, is not interrupted in CREM knockout mice [11]. Therefore, transcription factors other than A-myb and CREM are involved in early spermatid gene expression.
Several testis-specific transcripts whose expression is initiated in round spermatids have been characterized in the mouse and rat [40]. However, no information on their promoter analyses is available. Identification of promoter regions of additional testis-specific genes that drive round spermatid-specific gene expression in vivo, as was done in the present study, will help identify common cis-regulatory elements involved in coordinate gene transcription during early spermiogenesis. One interesting observation that has emerged from the promoter analysis of the mSP-10 gene is that a relatively short proximal promoter piece (294 bp) is sufficient to drive testis-specific gene expression in vivo. Short proximal promoter regions of other testis-specific genes, including protamine 1 (119 bp), t-ACE (91 bp), proenkephalin (119 bp), Pdha-2 (187 bp), and Ldhc-4 (100 bp), were also shown to be adequate for germ line-specific activity in transgenic mice [5, 6, 38, 41, 42], indicating that core promoter structure might be critical for testis-specific gene activation.
This is the first report on identification of a testis-specific gene promoter that activates round spermatid-specific gene expression. We are in the process of dissecting the -266 to +28-bp SP-10 promoter region further to determine cis-elements that interact with nuclear proteins from testis. The SP-10 gene promoter identified here thus may provide a useful tool with which to search for germ cell type-specific transcription factors. A central issue for the future will be to determine whether a common mechanism controls the transcriptional program for round spermatid growth and differentiation. Moreover, understanding the molecular basis of gene expression during acrosomal biogenesis, which marks the beginning of the terminal phase of male germ cell differentiation, may provide additional insight into defects in spermatogenesis leading to azoospermia and infertility.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: P. Prabhakara Reddi, Department of Cell Biology, Box 439, School of Medicine, University of Virginia, 1300 JPA, Charlottesville, VA 22908. FAX: 804 982 3912; ppr5s{at}virginia.edu
Accepted: June 17, 1999.
Received: April 26, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, a germ cell specific transcription factor involved in the coordinate expression of the zona pellucida genes. Development 1997; 124:4939–4947.[Abstract]This article has been cited by other articles:
|
|
 |
|
  M. Akerfelt, E. Henriksson, A. Laiho, A. Vihervaara, K. Rautoma, N. Kotaja, and L. Sistonen Promoter ChIP-chip analysis in mouse testis reveals Y chromosome occupancy by HSF2 PNAS, August 12, 2008; 105(32): 11224 - 11229. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Ranpura, U. Deshmukh, and P. P. Reddi NF45 and NF90 in Murine Seminiferous Epithelium: Potential Role in SP-10 Gene Transcription J Androl, March 1, 2008; 29(2): 186 - 197. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Tang, A. Kung, and E. Goldberg Regulation of Murine Lactate Dehydrogenase C (Ldhc) Gene Expression Biol Reprod, March 1, 2008; 78(3): 455 - 461. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M. Abhyankar, C. Urekar, and P. P. Reddi A Novel CpG-free Vertebrate Insulator Silences the Testis-specific SP-10 Gene in Somatic Tissues: ROLE FOR TDP-43 IN INSULATOR FUNCTION J. Biol. Chem., December 14, 2007; 282(50): 36143 - 36154. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Hsia, J. P. Brousal, S. R. Hann, and G. A. Cornwall Recapitulation of Germ Cell- and Pituitary-Specific Expression With 1.6 kb of the Cystatin-Related Epididymal Spermatogenic (Cres) Gene Promoter in Transgenic Mice J Androl, March 1, 2005; 26(2): 249 - 257. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Yang, S. Medvedev, P. P. Reddi, R. M. Schultz, and N. B. Hecht The DNA/RNA-binding protein MSY2 marks specific transcripts for cytoplasmic storage in mouse male germ cells PNAS, February 1, 2005; 102(5): 1513 - 1518. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Siep, E. Sleddens-Linkels, S. Mulders, H. van Eenennaam, E. Wassenaar, W. A. Van Cappellen, J. Hoogerbrugge, J. A. Grootegoed, and W. M. Baarends Basic helix-loop-helix transcription factor Tcfl5 interacts with the Calmegin gene promoter in mouse spermatogenesis Nucleic Acids Res., December 7, 2004; 32(21): 6425 - 6436. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L.-X. Feng, Y. Chen, L. Dettin, R. A. R. Pera, J. C. Herr, E. Goldberg, and M. Dym Generation and in Vitro Differentiation of a Spermatogonial Cell Line Science, July 19, 2002; 297(5580): 392 - 395. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. M. Eddy Male Germ Cell Gene Expression Recent Prog. Horm. Res., January 1, 2002; 57(1): 103 - 128. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. P. Lahti, R. Shariatmadari, J. K. Penttinen, J. R. Drevet, B. Haendler, M. Vierula, M. Parvinen, I. T. Huhtaniemi, and M. Poutanen Evaluation of the 5'-Flanking Regions of Murine Glutathione Peroxidase Five and Cysteine-Rich Secretory Protein-1 Genes for Directing Transgene Expression in Mouse Epididymis Biol Reprod, April 1, 2001; 64(4): 1115 - 1121. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
