| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Hubrecht Laboratory, Netherlands Institute for Developmental Biology (NIOB), 3584 CT Utrecht, The Netherlands
b Istituto di Anatomia degli Animali Domestici, Milan, Italy
c Department of Herd Health and Reproduction, Faculty of Veterinary Medicine, University of Utrecht, 3584 CL Utrecht, The Netherlands
d Laboratory of Immunogenetics, Department of Animal Sciences, University of Illinois at Urbana-Champaign, Champaign, Illinois 61820
e Laboratorio di Tecnologie della Riproduzione Consorzio per l'Incremento Zootecnico dell'Associazione Italiana Allevatori, 26100 Cremona, Italy
f Virologie Immunologie Moléculaires, Institut National de la Recherche Agronomiques, 78352 Jouy-en-Josas, France
g Institute of Reproduction and Development, Monash University, Monash Medical Centre, Clayton, Victoria, Australia
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the mouse, Pou5f1 is the earliest expressed gene known to encode a transcription factor, and in vivo its protein is present at low levels in unfertilized oocytes and is localized predominantly in the nuclei of preimplantation embryos at all cleavage stages. After differentiation of the trophectoderm lineage from the late blastocyst, Oct-4 is found only in inner cell mass (ICM) cells [6]. In direct contrast to earlier RNA localization studies [7, 8], Oct-4 was found at high levels in primitive endoderm cells that differentiate and migrate along the inner surface of the trophectoderm [6]. In the postimplantation mouse embryo, Pou5f1 transcripts were detected in embryonic ectoderm at Day 7 but not in endoderm, allantois, and other extraembryonic tissues or in mesoderm forming at Day 7.5 [7]. Pou5f1 mRNA in the mouse was further detected in male and female primordial germ cells after Day 8.5 of gestation but was absent in sperm and the testis, suggesting that its expression becomes restricted to the female germline at later stages of germ cell development [9].
Consistent with its embryonic expression pattern in vivo, Oct-4 is expressed in vitro in undifferentiated mouse embryonic stem (ES) cells and in mouse and human EC cells [3, 10], but it is strongly down-regulated during retinoic acid (RA)-induced differentiation in (murine) P19 and F9 EC cells [3, 11]. Analysis of the transcriptional control region by deletion mutagenesis revealed the presence of a cis RA-responsive element located between nucleotides -1132 and -889, which is composed of two domains (1A and 1B) and functions as an RA-repressible enhancer [12].
A second, more distantly located transcriptional control element was identified by transfection of a series of constructs composed of mouse Pou5F1 upstream sequences coupled to a LacZ reporter gene in mouse ES, EC, and embryonic germ (EG) cells [13]. This distal enhancer (DE) was shown to be critical for expression in ES and EG cells, although the level of activation was lower in EG than in ES cells. Consistent with the results of Okazawa and coworkers [12], these studies showed that expression in EC cells was controlled by a proximal enhancer (PE), likely to be the RA-responsive element located between nucleotides -1132 and -889. The PE was dispensable for Pou5f1 expression in ES and EG cells [13]. A third regulatory element of approximately 230 base pairs (bp) in the promoter region immediately upstream of the initiation codon was shown to be essential for expression in ES and EC cells (and presumably also EG cells), as deletion of this fragment abolished expression completely [13, 14]. Consistent with this regulation in vitro, Pou5f1 expression in preimplantation mouse embryos was dependent on the DE or germline enhancer with no reduction in LacZ expression upon deletion of the PE, whereas postimplantation expression was driven by the PE or epiblast enhancer [13]. Footprint analysis in EC and ES cells provided refinements of the location of critical regulatory elements in the promoter region to the Sp1 site (between nucleotides -52 and -42), domain 1A of the PE (between -1120 and -1100), and a site named 2A in the DE (between -2027 and -2010) which is similar to 1A of the PE, although situated in opposite orientation [15].
Despite extensive knowledge of expression and regulation of Pou5f1 in the mouse, to date only its human ortholog has been cloned and characterized [16]. No reports have been published describing cloning of POU5F1 orthologs in farm animal species despite its obvious importance in early embryonic development, probably at the top of a cascade of regulatory genes that induce and regulate differentiation, and despite relatively intensive but as yet only partially successful attempts to isolate pluripotent stem cells [17] or ES cells (reviewed in [18]). Especially in the light of the long generation interval of farm animals in comparison with mice and the high costs associated with in vivo validation of putative ES cells for germline transmission in these species, there is a need for (molecular) markers for a first screen for the pluripotency of such cells in vitro. The objective of the present study was to isolate the bovine ortholog of POU5F1 and evaluate its potential as a stem cell marker by following expression during early embryonic development.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A bovine genomic DNA library constructed in the BamHI sites of phage EMBL3 SP6/T7 (Clontech, Palo Alto, CA) was screened according to standard procedures [19] using a full-length mouse Pou5f1 cDNA clone [4] as probe. After three rounds of screening at increasing stringency, a single phage clone (
bPOU5F1) with an approximately 11-kilobase (kb) insert was isolated, which contained the entire coding sequence and intervening introns of bPOU5F1 as well as 2.9 kb of upstream sequence.
Restriction Mapping, Subcloning, andSequencing of bPOU5F1
A restriction map of bPOU5F1 was produced by (double) digestion with restriction endonucleases BamHI and XbaI (Gibco-BRL, Gaithersburg, MD). Restriction fragments harboring bPOU5F1 coding sequences were identified after probing the Southern blot with the full-length mouse cDNA and cloned in appropriately digested and dephosphorylated pBluescript SK- (Stratagene Inc., La Jolla, CA) or pGEM-7Zf(-) (Promega, Madison, WI) vector. Plasmid DNA suitable for automated sequencing was purified using a commercial miniprep kit (Qiagen, Chatsworth, CA). Subclones containing coding regions and upstream regulatory sequences were completely sequenced in both directions by primer walking using an automated nucleotide sequencer (ABI Systems, Foster City, CA). The resulting sequences were aligned and analyzed using IGsuite software (Intelligenetics, Leek, Belgium), LASERGENE software (DNASTAR Inc., Madison, WI), and the online Basic Local Alignment Search Tool (BLAST; [20]).
Transient Transfection of bPOU5F1 Reporter Constructs
For construction of bPOU5F1 luciferase (Luc) reporter constructs, first a blunt-end AvaI fragment (-1710/+48, Fig. 1a) was cloned in pBluescript SK-. A BamHI-XbaI (-1021/+ 48) fragment from this clone and a BamHI-XbaI fragment (-2892/-1021, Fig. 1a) were subsequently cloned in both orientations in the BamHI site of a promoterless reporter pLuc. Digestion of the sense bPOU5F1-Luc construct with StuI followed by ligation resulted in the construct 
PE (-1467/-889, Fig. 1a). DE,PE was made by ligating the BamHI-XbaI-digested bPOU5F1-Luc reporter. The bPOU5F1 proximal promoter construct was made by cloning the SauIIIA (-156/+20) fragment in the BamHI site of pLuc. Cell culture, transient transfections, and ß-galactosidase assays were performed as described before [21].
|
Genetic Mapping of bPOU5F1
Nine bulls that sire paternal half-sib families collectively known as the Illinois Reference and Resource Families (IRRF; [22]) were screened for a polymorphism in bPOU5F1 by single-strand conformational polymorphism (SSCP) analysis as described [23]. Polymerase chain reaction (PCR) products for SSCP analysis were generated using oligonucleotide primers OCT27 (5'-GGAAATTGGGAACACAATGGG-3', coding strand) and OCT28 (5'-AACAGCAACCTTCGTTTCGG-3', noncoding strand), which are both located downstream of the predicted stop codon of bPOU5F1 (see Fig. 1b). Offspring of heterozygous sires were subjected to SSCP analysis, and dams were included if available. Pairwise and multipoint linkage analysis of bPOU5F1 genotypes with IRRF markers across the genome [24] were performed as described [25].
RNA Isolation and Northern Blot Analysis of bPOU5F1
Steady-state RNA levels of bPOU5F1 in somatic tissues were determined on a Northern blot containing 10 µg of mRNA isolated from tissues collected from a still-born calf (brain, colon, heart, kidney, lung, and small intestine) or from adult animals (liver, ovary, placenta, testis) as described before [25]. The blot was hybridized essentially as described [26] with a 2127-bp BamHI-XbaI insert from a bPOU5F1 plasmid subclone containing exons 2, 3, 4, and 5 of the coding region (Fig. 1b). A separate Northern blot containing 10 µg of mRNA from undifferentiated and 10-6 M retinoic acid (RA)-treated Tera2cl13 human EC cells [27] was included as a control for hybridization. Blots were subsequently hybridized with a human glyceraldehyde 3-phosphate dehydrogenase (GAPDH) cDNA probe as a control for RNA loading.
Reverse Transcription (RT)-PCR of bPOU5F1 inOocytes and Preattachment-stage Bovine Embryos
Bovine ovaries were collected at a local abattoir in a thermoflask and transported to the laboratory. Cumulus oocyte complexes were obtained by aspiration of 2- to 8-mm follicles and selected on the basis of the presence of a multilayered cumulus investment. In vitro maturation and in vitro fertilization (IVF) of the cumulus oocyte complexes as well as in vitro embryo culture of the fertilized oocytes was performed as previously described [28]. Under the culture conditions employed, approximately 25% of fertilized oocytes are at the blastocyst stage at Day 9 postfertilization. Pools of 20 immature denuded oocytes (germinal vesicle [GV] stage) or embryos at the 2-, 4-, or 8-cell, morula, or Day 9 blastocyst stages of development, and one Day 14 and one Day 16 blastocyst were collected in 200 µl Ultraspec solution (Biotecx, Veenendaal, The Netherlands). ICM were obtained from Day 8 blastocysts produced in vitro as described [29], prepared by immunosurgery as described previously for mice [30], and collected individually in 200 µl Ultraspec solution. Total RNA was isolated as described previously in detail for human oocytes and individual preimplantation embryos [31]. RNA was treated with ribonuclease (RNase)-free deoxyribonuclease (DNase), dissolved in 10 µl RNase-free water, and reverse-transcribed in a 20-µl volume as described [31], with the exception that 1 µl of 0.5 mg/ml oligo dT12–18 (Pharmacia, San Francisco, CA) was used as primer. Two microliters of cDNA (4 µl in case of the Day 14 blastocyst) was used as template for amplification of bPOU5F1 in two rounds of PCR with hemi-nested primers.Primers for the first round were OCT15 and OCT26 (Table 1), which were both designed across splice junctions to favor amplification of cDNA over genomic DNA. OCT15 was designed from the murine nucleotide sequence [32] and contains a primer:template C:G mismatch at its most 5' base. Second-round PCR was performed using primers OCT26 as above and OCT29 (Table 1), which was modified at the 3' base to introduce an EagI restriction site into PCR products derived from bPOU5F1 but not into products derived from a bPOU5F1-related sequence (see Results). Reactions were carried out as described [31] with the exception that 0.6 units of Goldstar polymerase (Eurogentec, Seraing, Belgium) was used. This polymerase lacks 3'–5' exonuclease activity. The thermal cycling profile for the first round was initial denaturation for 5 min at 94°C, followed by 40 cycles of 15 sec at 94°C, 30 sec at 55°C, and 45 sec at 72°C. Final extension was for 7 min at 72°C. Subsequently, 2 µl of first-round product was transferred and amplified for 30 cycles according to the same profile, except that annealing was at 60°C. Water controls and RNA samples that had not been reverse-transcribed were included and transferred as described to monitor contamination and amplification from genomic DNA. Five microliters of second-round product was resolved and visualized as described [31], and another 5-µl aliquot of second-round PCR product was digested with 5 U EagI (New England Biolabs, Beverly, MA) for 1 or 16 h at 37°C in a 15-µl volume before gel electrophoresis to confirm the genuine origin from bPOU5F1. RNA isolations and RT-PCR were performed in duplicate for all stages examined. Genomic DNA was amplified in a single round of PCR as described using primers OCT15 and OCT12 (Table 1) to confirm the existence of bPOU5F1-related sequences.
|
Polyclonal Oct-4 Antiserum
Polyclonal antisera were raised against peptide NH2-NNENLQEICKSETLVQ-COOH corresponding to the linker region connecting the POU-specific and POU-homeodomains of Pou5f1 [3–5, 32]. The murine and bovine protein sequences in the region of the peptide are identical with the exception of a substitution of serine at position 216 in the mouse for alanine at the corresponding position 223 in cattle (Figs. 1b and 2). Peptide synthesis and (repeated) immunization of New Zealand White rabbits were performed as described [33]. Affinity-purified antibody was prepared by coupling the peptide to CnBr-activated Sepharose 4B by standard procedures (Pharmacia Biotech, St. Quentin-Yvelines, France), equilibrating the gel with PBS, loading the antiserum, and eluting peptide binding antibodies at acidic pH in 100 mM acetic acid, 300 mM NaCl, pH 3.0. The specificity of the affinity-purified fraction was determined by indirect immunofluorescence on P19 EC cells that were either undifferentiated or treated with 10-6 M RA for 5 days and on STO murine fibroblast cells (American Type Culture Collection).
|
Immunocytochemical Localization of Oct-4 in Bovine Oocytes, Preattachment-stage Embryos, andBlastocyst Outgrowths
Whole-mount immunocytochemistry was performed on bovine oocytes, embryos, and blastocyst outgrowths. The following developmental stages were examined: GV-stage oocytes; metaphase-II (M-II)-stage oocytes; pronuclear-stage zygotes; 2-cell, 4-cell, 8-cell, and 16-cell embryos; morulae; and early, expanded, hatched, and elongated blastocysts at Days 10, 14, and 16 postfertilization. Elongated blastocysts were obtained from in vitro-produced 2- to 4-cell-stage embryos [34]. For the generation of Day 10 blastocysts, embryos at the 2- to 4-cell stage were transferred in sheep oviducts and recovered 8 days later. To obtain Day 14 and Day 16 blastocysts, 2- to 4-cell-stage embryos were transferred in ligated sheep oviducts and recovered on Day 7 post-IVF. Embryos developed to the blastocyst stage were then transferred into the uterus of synchronized cows and recovered on Day 14 and Day 16 postestrus [29]. In order to determine Oct-4 expression in blastocyst outgrowths, in vitro-fertilized oocytes were cultured until the blastocyst stage at Day 8 on oviduct epithelial cells, transferred to buffalo rat liver-conditioned medium [35], and allowed to attach on gelatin-coated coverslips (Sarstedt, Numbrecht, Germany). At Day 13 postfertilization, four outgrowths were fixed and processed as for the other specimens.
A minimum of 10 specimens for each developmental stage were fixed at room temperature for 20 min in 2.5% (w:v) paraformaldehyde in PBS (pH 7.4), and for an additional 10 min in the same fixative with the addition of 0.3% (v:v) Triton X-100. Fixation was followed by an incubation in 50 mM ammonium chloride for 60 min to remove remaining free aldehyde groups. Samples were incubated with the primary antibody diluted 1:10 overnight at 4°C, followed by fluorescein isothiocyanate (FITC)-conjugated donkey anti-rabbit IgG (Jackson ImmunoResearch Labs., West Grove, PA) secondary antibody. After several washes, samples were mounted on slides in Vectashield Mounting Medium (Vector Labs., Burlingame, CA). Controls were performed using the primary antibody neutralized with a 10-fold excess of the peptide or by omission of the primary antibody. Oocytes, embryos, and outgrowths were examined and photographed with a Nikon Diaphot TMD fluorescence microscope (Nikon Instruments, Garden City, NY).
Confocal Laser Scanning Microscopy
Immunofluorescence localization of Oct-4 in expanded and hatched blastocysts was observed using a Meridian InSight Point Confocal Laser Scanning Microscope (CLSM) assembled on a Zeiss (Oberkochen, Germany) Axioscope microscope equipped with Zeiss ICS oil immersion lenses. FITC samples were observed using an argon laser with an excitation filter at 488–10 nm, a dichroic mirror at 505 nm, and a barrier filter LP 515 nm. Images of serial optical sections were recorded every 1 µm along the z-axis of each embryo. Three-dimensional reconstructions were made from each series of optical sections using the computer work station and the software provided by the manufacturer.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
bPOU5F1 with an insert of approximately 11 kb harboring the entire bovine POU5F1 gene was isolated by screening of a genomic DNA library with a full-length murine Pou5f1 cDNA clone as a probe. Two fragments were subcloned and sequenced in both directions: a 3721-bp BamHI fragment containing exon 1 (405-bp coding sequence), 2955 bp of promoter and upstream regulatory sequence, and 361 bp of the intron between exons 1 and 2 (Fig. 1a); and a 2127-bp BamHI-XbaI fragment containing exons 2 (121 bp), 3 (131 bp), 4 (159 bp), and 5 (267-bp coding sequence) and flanking sequences (Fig. 1b). The numbering of bovine exons is according to that of Pou5f1 [12], and the boundaries of exons 2–5 conform with the GT/AG consensus sequence of splice donor/acceptor sites [36]. The splice donor site following exon 1 is GC instead of GT (Fig. 1a), which is the most commonly found alternative splice donor sequence [37] and is also present at the corresponding position in the human POU5F1 ortholog [16]. The predicted Oct-4 protein encoded by these five exons consists of 360 amino acids (Fig. 2) and has 90.6% and 81.7% overall identity with its human [16] and murine [3–5, 32] counterparts, respectively. Broken down by exon, the percentages of amino acid identity between bovine and human Oct-4 are 84.4, 97.5, 100, 94.3, and 89.7 for exons 1 through 5, respectively. For the comparison between cattle and mice, these percentages are 71.0, 95.1, 95.4, 84.9, and 82.8, respectively. Alignment of the promoter sequence with the orthologous human [16] and mouse [12] sequences revealed extensive conservation in the region upstream of the transcription initiation site: the putative Sp1 binding site and overlapping hormone responsive element at nucleotides -56/-31 in the bovine sequence are identical in all three species (Fig. 3a). In addition, there is complete sequence conservation between the bovine and mouse sequence at sites that showed protection or hypermethylation in part 1A of the PE (Fig. 3b) and strong conservation at those we believe are the bovine equivalent of the DE (Fig. 3c). The corresponding sequence in humans at these positions is not known. Alignments of the bovine and murine upstream sequences further indicated several other regions of strong homology, of which a 171-bp region defined by the bovine sequence at nucleotides -1270/-1100 (Fig. 1a) with 96% identity is most significant. This region contains part 1B of the PE as defined by Okazawa and coworkers [12]. Another 13-bp region, 5'-GGGGGAGGGAGAA-3', at -1037/-1024 is 100% conserved with a mouse sequence downstream of the PE (Fig. 1a) and closely resembles the critical region of part 1A of the PE (Fig. 3b). BLAST analysis (data not shown) revealed a 183-bp region with 86% identity to the bovine Short Interspersed Nuclear Element (SINE) Bov-tA consensus sequence [38] at positions -697/-517 (Fig. 1a), which is absent in the mouse.
|
Transient transfection experiments in P19 EC cells using luciferase (Luc) reporter constructs were performed to determine whether the bPOU5F1 promoter is functional and harbors regulatory elements mediating down-regulation in response to RA. Highest Luc activity (average 14 500 arbitrary units) was observed when the entire promoter and upstream sequences (-2892/+48) were cloned in correct orientation upstream of the Luc gene, which was down-regulated to basal level upon treatment with RA. Control transfections of constructs with bPOU5F1 promoter and regulatory elements in opposite orientation showed much less Luc activity (average 3800 units). Deletion of the putative PE resulted in reduced Luc expression in comparison with the full-length construct (average 5500 units), but this activity was still repressible by RA (average 1088 units), demonstrating that the PE functions as an enhancer but that down-regulation in response to RA also occurs without the PE. Deletion of the PE and DE together showed low levels of Luc activity (average 1500 units) comparable to that of a construct with only the proximal (-150/+48) bPOU5F1 promoter, and was independent of treatment with RA (Fig. 4).
|
The genetic localization of bPOU5F1 was determined by segregation analysis of an SSCP in a 257-bp PCR product amplified from the region downstream of the predicted stop codon of the gene. Linkage analysis performed on the genotypes of offspring from three heterozygous IRRF sires placed bPOU5F1 on bovine chromosome 23 in the region of the major histocompatibility complex (MHC), also known as the bovine leukocyte antigen (BoLA) complex. No recombinants were observed between bPOU5F1 and the gene for 21-steroid hydroxylase (CYP21), the MHC class I locus BoLA-A, and microsatellite marker TAMLS113.3, with more than 30 informative meioses for each pair of loci (Table 2; [24]). As a result, the order of bPOU5F1 with respect to these three closely linked loci could not be resolved. However, multipoint analysis (data not shown) resulted in the order (centromere) - BoLA-DRB3 (recombination fraction (
) = 0.01; (log likelihood ratio (Llhr) = 6.62) - [bPOU5F1, BoLA-A, CYP21, TAMLS113.3] ( = 0.01; Llhr = 0.39) - MGTG7 ( = 0.05; Llhr = 5.10) - [RM185, prolactin (PRL)] - (tel) as most likely, which places bPOU5F1 with strong statistical confidence between BoLA-DRB3 and PRL.
|
Steady-state RNA levels of bPOU5F1 in tissues from a newborn calf and adult animals was determined by Northern blot analysis using an homologous probe. No specific signal was detected in colon, small intestine, heart, brain, liver, lung, kidney, ovary, placenta, and testis tissues after 5 days of exposure (Fig. 5a) or after 12 days of exposure (data not shown). By contrast, POU5F1 mRNA could be readily detected in undifferentiated Tera2 cells on a separate blot that was hybridized simultaneously and exposed for 2 days (Fig. 5c). GAPDH mRNA was detected in all bovine samples (Fig. 5b). These results suggest that there are very low levels of bPOU5F1 mRNA present in these tissues or none at all.
|
The presence of bPOU5F1 transcripts in oocytes and during preattachment development was determined by RT-PCR. Control PCR experiments using primers OCT12 and OCT15 and bovine genomic DNA as a template resulted in PCR products in the approximate size range of the 333 bp expected from bPOU5F1-derived cDNA, indicative of the existence of related (pseudo)genes (data not shown). Cloning of these products and sequencing of twelve randomly chosen clones revealed two distinct sequences, bPOU5F1rs1 and bPOU5F1rs2, with 99.3% and 72.9% nucleotide identity, respectively, to the corresponding coding regions of bPOU5F1 (Fig. 6). On the basis of this observation and the assumption that these bPOU5F1-related sequences could be transcribed, an RT-PCR assay specific for bPOU5F1 was developed; primer OCT12 was omitted to eliminate amplification of bPOU5F1rs2, and primer OCT29 was designed to introduce an EagI restriction site in PCR products derived from bPOU5F1 but not bPOU5F1rs1, because of a nucleotide difference between these sequences at position 186 of the alignment depicted in Figure 6. Using this RT-PCR assay, specific amplification products were obtained from cDNA but not from control RNA of GV-stage oocytes and in vitro-fertilized and cultured 2-, 4-, and 8-cell-, morula-, and Day 9 blastocyst-stage embryos, ICM cells isolated from a Day 8 blastocyst, and Tera2cl13 cells as control (Fig. 7A). Digestion of these PCR products with EagI resulted in partial cleavage in all samples (Fig. 7B), indicative of the presence of bPOU5F1 transcripts. The origin of undigested PCR products that were also observed in all cDNA samples but not in RNA controls is not clear. The results of duplicate experiments were consistent for all stages examined. No amplification products were obtained from Day 14 and Day 16 blastocysts or from water controls (data not shown).
|
|
Immunocytochemical evaluation of Oct-4 expression in oocytes and embryos revealed expression in the cytoplasm and the nucleus of all stages examined until Day 10 of development (Figs. 8 and 9). Evaluation of stained specimens showed a marked nuclear localization of Oct-4 during segmentation but a more diffuse distribution upon formation of the blastocoel (Fig. 8). Oct-4 was detected with CLSM in ICM and trophectoderm cells of expanded and hatched blastocysts at Days 8, 9, and 10 after fertilization (Figs. 9 and 10). Oct-4 protein could not be detected in blastocysts at Days 14 (Fig. 10) and 16 (data not shown) of development. Oct-4 staining was visible in putative ICM cells and cells attached to the culture surface when blastocyst outgrowths were analyzed (Fig. 11). Control experiments resulted in nuclear staining of almost all undifferentiated P19 EC cells but not of somatic cells (data not shown). No staining was observed in control experiments using Oct-4 antibody preincubated with excess peptide or omitting the primary antibody (data not shown).
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Expression of bPOU5F1 at the RNA and protein levels was observed in bovine oocytes and embryos until Day 10 of development, consistent with a role in early embryo development, as has been postulated in the mouse [4, 5, 7, 8]. The most notable difference with expression of Oct-4 during murine preimplantation development [6] is that in cattle Oct-4 is detected at the (late) blastocyst stage in ICM and trophoblast cells; this is followed by coordinate down-regulation in these cell types around Day 10 postfertilization with no detectable expression after Day 12. Expression of Oct-4 in the bovine is thus not limited to pluripotent cells of the early embryo, which warrants careful interpretation when it is used as a marker for pluripotency of cultured (embryonic) cells.
Analysis of promoter and transcriptional control regions of bPOU5F1 and transient transfections of Luc reporter constructs in P19 EC cells showed strong conservation of sequence and function with the orthologous regions in the mouse, especially at the proximal promoter and PE. The only exception is a bovine SINE element in bPOU5F1. It will be interesting to determine whether the different expression patterns of Oct-4 in trophoblast cells of cattle and mice are mediated by transcriptional regulatory elements or rather reflect other species-specific factors influencing bPOU5F1 transcription or protein clearance. This hypothesis can be tested by microinjection of bovine reporter constructs in mouse embryos as described [13] and determining whether the expression follows that of the mouse or the cow.
McWhir and coworkers [41] described a method for derivation of ES cell lines from murine embryos with a nonpermissive genetic background using selective ablation of differentiating cells. In order to determine whether a similar approach might be feasible in cattle, the outgrowths formed at Day 13 from in vitro-produced Day 9 blastocysts were examined for expression of Oct-4 by immunofluorescence. Our observation of Oct-4 expression in all cells, including two morphologically distinct cell types that may account for the ICM with its immediate differentiated derivatives [42] and trophoblast cells, suggests that apart from the difficulty of producing transgenic animals, this approach might be more complicated in cattle than in mice. However, additional experiments with prolonged culture of blastocyst outgrowths or isolated ICMs, or the identification of independent specific markers for undifferentiated bovine embryonic cells, are needed to investigate this in more detail. Interesting in this respect is that alkaline phosphatase (AP) activity, which in the mouse is a marker for undifferentiated cells in vivo and in vitro, is also expressed in ICM and trophoblast cells of the bovine blastocyst but showed a heterogeneous staining pattern in plated ICM cells cultured for 2–4 days [42], suggesting that either bovine ICM cells rapidly differentiate in culture or that AP is not restricted to pluripotent cells in bovine embryos. Oct-4 thus seems to behave rather as AP in the early bovine embryo.
Brook and Gardner [43] presented an alternative strategy for the derivation of mouse ES cell lines by explanting blastocyst tissues that had been microsurgically isolated from the permissive 129 strain and several hitherto nonpermissive strains of mice. This study showed that ES lines originate from the epiblast and could be established with high efficiency when using epiblasts isolated from Day 5 blastocysts, either intact or as a single cell suspension. Efficiencies of ES line derivation even reached 100% when epiblasts from Day 9 delayed-implantation 129 strain blastocysts were used. The authors claimed that their method is superior to that described by McWhir and coworkers [41] on the basis of the high success rate, the fact that there was no question about the genetic background of the mouse strains involved, and the fact that the neo gene could still be used as selection marker of transfected cells once lines had been derived; they suggested that their approach may be applicable to other species. Our observation of a broader expression pattern of Oct-4 during bovine early embryonic development than in the mouse, together with the compelling results achieved by Brook and Gardner [43] argue that further attempts to derive pluripotent cell lines or ES lines in cattle should be directed at mimicking the latter approach using (isolated) ICM cells of preattachment embryos at different stages of development. A similar approach has earlier led to the birth of identical offspring following nuclear transfer from cultured bovine ICM cells but not to proven ES lines [17].
Cibelli and coworkers [44] recently reported the production of transgenic chimeric offspring from bovine ES-like cells derived from blastocysts resulting from nuclear transfer with transgenic fetal fibroblast cells. These ES-like cells, which could not be clonally propagated, lacked the differentiation markers vimentin and cytokeratin and were AP-negative. They were not investigated for Oct-4 expression. Although transgenic chimerism was demonstrated, including in gonadal tissues, the authors refer to these cells as ES-like cells because germline transmission was not demonstrated. It therefore still remains to be seen to what extent differences in early embryological development between mice and cattle, which do not have a discrete epiblast, as well as culture conditions employed, will prove a surpassable barrier in establishing ES lines from early bovine embryos.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: F. Gandolfi, Istituto di Anatomia degli Animali Domestici, Via D. Trentacoste 2, Milan, Italy. FAX 3902 214 0745; fulvio.gandolfi{at}unimi.it
Accepted: December 15, 1998.
Received: June 16, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. I Alexopoulos, P. Maddox-Hyttel, P. Tveden-Nyborg, N. T D'Cruz, T. R Tecirlioglu, M. A Cooney, K. Schauser, M. K Holland, and A. J French Developmental disparity between in vitro-produced and somatic cell nuclear transfer bovine days 14 and 21 embryos: implications for embryonic loss Reproduction, October 1, 2008; 136(4): 433 - 445. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Wuensch, F. A. Habermann, S. Kurosaka, R. Klose, V. Zakhartchenko, H.-D. Reichenbach, F. Sinowatz, K. J. McLaughlin, and E. Wolf Quantitative Monitoring of Pluripotency Gene Activation after Somatic Cloning in Cattle Biol Reprod, June 1, 2007; 76(6): 983 - 991. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. Misica-Turner, F. C. Oback, M. Eichenlaub, D. N. Wells, and B. Oback Aggregating Embryonic but Not Somatic Nuclear Transfer Embryos Increases Cloning Efficiency in Cattle Biol Reprod, February 1, 2007; 76(2): 268 - 278. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Pebay, R. C.B. Wong, S. M. Pitson, E. J. Wolvetang, G. S.-L. Peh, A. Filipczyk, K. L.L. Koh, I. Tellis, L. T.V. Nguyen, and M. F. Pera Essential Roles of Sphingosine-1-Phosphate and Platelet-Derived Growth Factor in the Maintenance of Human Embryonic Stem Cells Stem Cells, October 1, 2005; 23(10): 1541 - 1548. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. C. Chang, H. Liu, J. Zhang, J. Grifo, and L. C. Krey Developmental incompetency of denuded mouse oocytes undergoing maturation in vitro is ooplasmic in nature and is associated with aberrant Oct-4 expression Hum. Reprod., July 1, 2005; 20(7): 1958 - 1968. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Cauffman, H. Van de Velde, I. Liebaers, and A. Van Steirteghem Oct-4 mRNA and protein expression during human preimplantation development Mol. Hum. Reprod., March 1, 2005; 11(3): 173 - 181. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Vejlsted, B. Avery, M. Schmidt, T. Greve, N. Alexopoulos, and P. Maddox-Hyttel Ultrastructural and Immunohistochemical Characterization of the Bovine Epiblast Biol Reprod, March 1, 2005; 72(3): 678 - 686. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kurosaka, S. Eckardt, and K. J. McLaughlin Pluripotent Lineage Definition in Bovine Embryos by Oct4 Transcript Localization Biol Reprod, November 1, 2004; 71(5): 1578 - 1582. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
|
