Differences in early lineage segregation between mammals


  • Ewart W. Kuijk,

    1. Department of Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
    Search for more papers by this author
  • Leonie Du Puy,

    1. Department of Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
    2. Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
    Search for more papers by this author
  • Helena T.A. Van Tol,

    1. Department of Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
    Search for more papers by this author
  • Christine H.Y. Oei,

    1. Department of Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
    Search for more papers by this author
  • Henk P. Haagsman,

    1. Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
    Search for more papers by this author
  • Ben Colenbrander,

    1. Department of Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
    Search for more papers by this author
  • Bernard A.J. Roelen

    Corresponding author
    1. Department of Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
    • Department of Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, Yalelaan 104, 3584 CM Utrecht, The Netherlands
    Search for more papers by this author


Two lineage segregation events in mammalian development form the trophectoderm, primitive endoderm, and pluripotent primitive ectoderm. In mouse embryos, Oct4, Cdx2, Nanog, and Gata6 govern these events, but it is unknown whether this is conserved between mammals. Here, the expression patterns of these genes and their products were determined in porcine oocytes and embryos and in bovine embryos. CDX2 and GATA6 expression in porcine and bovine blastocysts resembled that of mouse, indicating conserved functions. However, NANOG expression was undetectable in porcine oocytes and embryos. Some inner cell mass cells in bovine blastocysts expressed NANOG protein. OCT4 protein was undetectable in porcine morulae, but present in both the trophectoderm and the inner cell mass of blastocysts, suggesting that downregulation of OCT4 in the trophectoderm does not precede trophectoderm formation. Combined, the results indicate differences in lineage segregation between mammals. Developmental Dynamics 237:918–927, 2008. © 2008 Wiley-Liss, Inc.


In the mouse, there is considerable knowledge on the molecular mechanisms of early development, with the transcription factors Oct4, Cdx2, Nanog, and Gata6 as key players. Two successive differentiation events in early embryonic development result in the segregation of three committed lineages that together form the blastocyst. The first segregation occurs at the compacted morula stage with the outer layer of cells forming the epithelial trophectoderm (TE), which becomes the embryonic part of the placenta, and the inner layer of cells forming the inner cell mass (ICM), which produces embryonic cells and the extra-embryonic mesoderm and primitive endoderm. Embryos that lack Oct4 fail to form an ICM that can differentiate along embryonic lineages and cells are restricted to differentiation along the extraembryonic trophoblast lineage (Nichols et al.,1998), whereas embryos that lack Cdx2 are unable to maintain the TE lineage (Strumpf et al.,2005). Furthermore, Oct4 and Cdx2 mutually inhibit each other's expression and both transcription factors are detected in all of the nuclei of 8-cell-stage embryos (Niwa et al.,2005). These findings have resulted in a model in which mutual repression of these transcription factors in early morulae results in Cdx2-expressing and Oct4-expressing cells, with loss of Cdx2 expression in the ICM as a primary event, and with segregation of the inner cell mass and trophectoderm lineages as a consequence (Niwa et al.,2005; Ralston and Rossant,2005; Strumpf et al.,2005).

The second segregation of lineages divides the inner cell mass into the primitive ectoderm, which gives rise to the embryo proper and the primitive endoderm (PE) that forms the extra-embryonic endoderm layer of the visceral yolk sac and in rodents also the parietal endoderm. Before the PE is formed, its precursors can already be detected by expression of Nanog and Gata6, which shows a so-called “pepper-and-salt” distribution, with Nanog-positive cells destined to become epiblast and Gata6-positive cells destined to become PE (Chazaud et al.,2006). The mosaic distribution of these precursors is considered to depend on Grb2-Ras-MAP kinase signaling, because inactivation of Grb2 results in Nanog expression in all cells of the ICM and loss of Gata6 expression (Chazaud et al.,2006). Embryonic stem (ES) cells in which Gata6 or a close family member Gata4 are over-expressed develop into PE cells (Fujikura et al.,2002), whereas ES cells that lack Gata6 or Gata4 fail to develop visceral endoderm in in vitro differentiation experiments such as embryoid body cultures (Morrisey et al.,1998; Koutsourakis et al.,1999). Mouse embryos that lack Nanog develop TE and PE, but fail to form an epiblast and loss of Nanog in ES cells results in extra-embryonic endoderm-like cells (Mitsui et al.,2003). As a consequence, the murine epiblast is defined by two segregation events that depend on expression of, respectively, Oct4 and Nanog, whereas Cdx2 specifies the trophectoderm, and Gata6 the primitive endoderm.

Stem cells can be derived from all three lineages (Ralston and Rossant,2005). Only ES cells derived from the ICM of blastocysts that express Oct4 (Niwa et al.,2000) and Nanog (Chambers et al.,2003) are pluripotent, which means they have the potential to become any specialized cell of all three embryonic germ layers even after prolonged culture. Therefore, these cells are important for differentiation studies and future regenerative medicine (Thomson et al.,1998). Thus the mechanism of early lineage segregation also defines the pluripotent cell population in blastocysts.

Recently, it was demonstrated that fibroblasts can be reprogrammed to a pluripotent state by retroviral induction of just four factors, that is Oct4, c-Myc, Klf4, and Sox2 (Takahashi and Yamanaka,2006). If these cells were selected for expression of Nanog, these induced pluripotent stem cells could colonize the germline, thereby unraveling an important part of the mechanism behind pluripotency (Okita et al.,2007; Wernig et al.,2007). Still, the efficiency of reprogramming was rather low and it is unclear why only some cells were reprogrammed to the pluripotent state. Interestingly, an oocyte has the same capacity to reprogram differentiated cells to the pluripotent state, as made evident by cloning through somatic cell nuclear transfer (Campbell et al.,1996). Therefore, it is important to study the dynamics of oocyte and early embryo development, to increase our knowledge on processes that contribute to pluripotency of cells.

Preimplantation development in mammals shows remarkable differences between species, possibly influencing the mechanism responsible for the formation of a pluripotent cell population. For instance, mouse embryos form an egg cylinder after implantation, whereas human, bovine, and porcine embryos have a planar morphology (Behringer et al.,2000). Furthermore, mouse and human embryos invasively implant at the blastocyst stage, which results in a haemochorial placenta. However, porcine and bovine blastocysts elongate before implantation, transforming from a sphere of a few millimeters in diameter to a long thin filament that in pigs can reach up to 100 cm in length at the time of implantation. This results in a loose diffuse non-invasive epitheliochorial placenta (Enders and Carter,2004). Moreover, OCT4 expression in bovine and porcine embryos is not limited to the ICM (van Eijk et al.,1999; Kirchhof et al.,2000), which suggests a difference in mechanism of the earliest lineage segregation between species. As a consequence, species differ in the factors that contribute to the establishment of the pluripotent cell population in embryos, which could explain why ES cell lines from species such as cow and pig have not been established yet (Keefer et al.,2007). Therefore, a better understanding of the mechanisms underlying pluripotency in these species is needed (Blomberg et al., 2008). In order to obtain more insight in early lineage segregation events and the establishment of the pluripotent cell population in these species, the expression patterns of NANOG, OCT4, CDX2, GATA4, and GATA6 were studied during early porcine and bovine embryonic development. The expression patterns of porcine and bovine orthologs of key genes in lineage segregation indicate diversity in early lineage segregation between mammals.


Gene Expression Patterns

Porcine oocytes and embryos with good morphology (Van Soom et al.,2003), ranging from germinal vesicle stage oocytes to blastocysts, were collected from three independent porcine in vitro cultures. These samples were used to study gene expression patterns using quantitative RT-PCR (QPCR) of genes important for pluripotency (NANOG and OCT4), for the development of the TE (CDX2), and for the development of the primitive endoderm (GATA4 and GATA6). PCR products were of the anticipated sizes (see Supplemental Fig. 1, which can be viewed at www.interscience.wiley.com/jpages/1058-8388/suppmat) and sequence analysis confirmed amplification of the desired product (data not shown).

In the case of NANOG, genomic DNA served as a positive control. All samples, except for one 4-cell-stage embryo, showed very late threshold cycles, which indicates low starting quantities. In 9 samples, expression levels did not exceed background levels from their corresponding -RTs, which surprisingly included all expanded blastocysts. This suggests that NANOG does not play an important role in porcine embryos at this time of development. Therefore, NANOG was excluded from the quantitative PCR analysis. For all other genes, expression levels were normalized to the geometric mean of GAPDH, PGK1, S18, and UBC. These reference genes allow direct comparison of gene expression levels in early porcine developmental stages ranging from oocytes to blastocysts (Kuijk et al.,2007).

OCT4 expression showed a 5-fold up-regulation from the germinal vesicle (GV) stage to the metaphase 2 (M2) stage (Fig. 1), indicating that OCT4 could be involved in oocyte maturation. At the 2-cell stage, OCT4 expression had dropped drastically and at the 4-cell stage, OCT4 expression was restored to GV stage levels. OCT4 expression was significantly higher in blastocysts than in cleavage stage embryos, which indicates a more prominent role for OCT4 at these stages. CDX2 expression was upregulated more than 10-fold in blastocysts when compared with GV stage oocytes.

Figure 1.

Relative expression of genes specific for the ICM, TE, or PE in porcine oocytes and preimplantation embryos. X-axis: Developmental stage. GV, germinal vesicle stage; M2, metaphase-2 stage; 2C, 2-cell stage; 4C, 4-cell stage; CB, early cavitating blastocyst; EB, expanded blastocyst. Y-axis: Normalized relative expression. For each developmental stage, the normalized expression value was divided by the normalized expression value of the germinal vesicle stage. Asterisks denote significant differences between the stages that are within the brackets. a denotes significant differences between GV stage and M2 stage oocytes. Error bars represent standard deviation. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Of those genes that in mouse embryos are important for PE formation, GATA4 expression showed a 4-fold higher expression in M2 stage oocytes than in GV stage oocytes, which suggests that GATA4 plays a role in oocyte maturation. Blastocysts showed significantly less GATA4 expression than cleavage stage embryos. GATA6 expression was more than 20-fold upregulated in blastocysts when compared with 4-cell embryos and expression of this gene was significantly higher in blastocysts than at earlier stages.

Additionally, the expression pattern of UTF1, which codes for a protein correlated with pluripotency in mouse embryos (Okuda et al.,1998; van den Boom et al.,2007) and the expression pattern of CK18, coding for an intermediate filament protein highly expressed in blastocysts (Brulet et al.,1980; Oshima et al.,1983), were investigated by QPCR in porcine oocytes and embryos (Fig. 1). UTF1 expression was comparatively low in both oocytes and cleavage stages and was significantly higher in both blastocyst stages, with more than six-fold up-regulation compared to GV stage oocytes (Fig. 1). This induction in expression correlates with previously observed specific expression of UTF1 in cells of the ICM of mouse blastocysts (Okuda et al.,1998). The expression of CK18 showed more than 10-fold higher expression in blastocyst stage embryos.

Whole Mount Immunofluorescence

Porcine embryos.

In vitro and in vivo produced porcine embryos were used in immunofluorescence studies to examine the expression and localization of factors associated with lineage segregation. Of those proteins that are involved in formation of pluripotent primitive ectoderm cells in the mouse, NANOG was neither detectable in in vitro produced porcine blastocysts (Fig. 2), nor in in vivo produced morulae and blastocysts (see Fig. 4). This is compatible with the QPCR data and suggests a lack of NANOG contribution to lineage segregation processes at these stages. In concordance with previous findings (Kirchhof et al.,2000) and in line with the gene expression levels, OCT4 was expressed in all cells of in vitro (Suppl. Fig. 2) and in vivo produced blastocysts, whereas in vivo produced morulae lacked OCT4 expression (Fig. 3).

Figure 2.

Immunofluorescence results on in vitro–produced porcine blastocysts. AC: Embryo, NANOG stain, and overlay of NANOG with DNA stain, respectively. DF: Embryo, CDX2 stain, and overlay of CDX2 with DNA stain, respectively. Dashed line denotes ICM. Arrow, CDX2 negative ICM cell. GI: Embryo, GATA4 stain, and overlay of GATA4 with DNA stain, respectively. Scale bar = 50 μm.

Figure 4.

Immunofluorescence results on in vitro–produced bovine blastocysts. AC: Blastocyst, NANOG stain, and overlay of NANOG with DNA stain, respectively; dashed line denotes ICM. Arrowhead denotes NANOG-negative ICM cell. DF: Blastocyst, GATA6 stain, and overlay of GATA6 with DNA stain, respectively; dashed line denotes ICM. Arrowhead denotes GATA6-negative ICM cell. GI: Blastocyst, GATA6 stain, and overlay of GATA6 with DNA stain, respectively; dashed line denotes ICM with GATA6-positive cells aligning the ICM. JL: Blastocyst, CDX2 stain, and overlay of CDX2 with DNA stain, respectively. Dashed line denotes CDX2-negative ICM. Scale bar = 50 μm.

Figure 3.

Immunofluorescence results on in vivo–produced porcine morulae and blastocysts. AC: Morula, NANOG stain, and overlay of NANOG with DNA stain, respectively. DF: Blastocyst, NANOG stain, and overlay of NANOG with DNA stain, respectively. GI: Morula, OCT4 stain, and overlay of OCT4 with DNA stain, respectively. Arrow denotes aspecific binding of antibody to zona pelucida. JL: Blastocyst, OCT4 stain, and overlay of OCT4 with DNA stain, respectively. Arrow denotes aspecific binding of antibody to zona pelucida. Arrowhead denotes OCT4-positive TE cell. MO: Blastocyst, GATA6 stain, and overlay of GATA6 with DNA stain, respectively. Arrowhead, GATA6-negative ICM cell. Dashed line denotes ICM. Scale bar = 50 μm.

Expression of CDX2, a transcription factor involved in the development of murine TE, was also restricted to TE cells of in vitro produced porcine embryos, which is in agreement with its mRNA levels and suggests that CDX2 contributes to the formation of porcine TE (Fig. 2).

Of those proteins that are associated with mouse PE, GATA4 was not detected in in vitro produced porcine blastocysts, whereas sections of paraffin-embedded testicular tissue, which served as a positive control, showed positive staining for GATA4 in Sertoli cells (Suppl. Fig. 3) (Ketola et al.,1999). This is in line with its mRNA expression and suggests GATA4 is not involved in PE formation in pigs. In contrast to mRNA expression levels, GATA6 protein could not be detected in in vitro–produced blastocysts (data not shown), which suggests that transcripts are not translated yet. On the other hand, some cells of the ICM of in vivo–produced blastocysts expressed GATA6 (Fig. 3), which indicates that this factor is involved in establishing porcine PE. All signals were detected above background levels of isotype controls (Suppl. Fig. 3).

Bovine embryos.

In order to determine whether the expression patterns of key players in murine early lineage segregation behave differently in other mammals, the spatial expression of these proteins was also studied in blastocysts of another ungulate species, the cow. As in mouse embryos but contrary to pig embryos, several cells of the ICM of in vitro–produced bovine blastocysts were positive for NANOG expression, while others were negative (Fig. 4), which indicates NANOG is expressed in cells of the ICM that will contribute to the primitive ectoderm. CDX2 was detected in TE cells of bovine blastocysts, but it was not detected in cells of the ICM, which suggests a role of this factor in the formation of the bovine TE. Also in line with mouse development, GATA6 displayed a mottled expression pattern in the ICM of bovine blastocysts (Fig. 4). Occasionally, GATA6-positive cells aligned the ICM in a PE-like fashion (Fig. 4). This suggests a role for GATA6 in PE development in cows. All signals were detected above background levels of isotype controls (Suppl. Fig. 4).


Most of our molecular knowledge on early embryonic development comes from studies on mouse embryos and ES cells. These studies have resulted in a model for the first two differentiation events: firstly, segregation of the TE from the ICM as a consequence of reciprocal inhibition of Oct4 and Cdx2 (Niwa et al.,2005; Ralston and Rossant,2005; Strumpf et al.,2005) and, secondly, Grb2-mediated mosaic expression of Nanog and Gata6 in the ICM, which causes subsequent segregation of the primitive ectoderm from the primitive endoderm (Chazaud et al.,2006). These events define the embryonic pluripotent cell population, which, in the case of mouse (Evans and Kaufman,1981) and human (Thomson et al.,1998), can be isolated and cultured without loss of pluripotency. The nucleus of a somatic cell can become pluripotent by transfer into an enucleated oocyte by which the genome is reprogrammed to a pluripotent state (Campbell et al.,1996). Knowledge on oocytes and early lineage segregation events will help to resolve the mechanism of pluripotency. Embryonic differences between mammals indicate that the pluripotent cell population is established differentially between species. For example, expression of OCT4 in TE of pigs and cows demonstrates that this factor is not involved in the segregation of TE and ICM in these species (van Eijk et al.,1999; Kirchhof et al.,2000).

In vitro fertilized oocytes are susceptible to polyspermy leading to abnormal embryo formation. In this study, sow oocytes were used instead of those from pre-pubertal gilts, in order to minimize the occurrence of polyspermy (O'Brien et al.,1996; Marchal et al.,2001). Of those factors that play a part in the formation of the primitive ectoderm, NANOG mRNA levels did not exceed background levels in in vitro–produced porcine embryos. In support of the mRNA expression levels in the current study, NANOG protein was also not detected in in vitro– and in vivo–produced porcine embryos, even though the antibody used in this study was able to detect porcine NANOG protein in another porcine tissue (unpublished data). Some cells of the ICM of bovine blastocysts expressed NANOG. A recent study described detection of NANOG mRNA in isolated ICM from in vivo–derived day-8 pig embryos (Blomberg et al., 2007). However, the authors used RNA amplification and the relative quantity was significantly less than in isolated ICMs after 24 hr of culture (Blomberg et al., 2008). In another study, low levels of NANOG mRNA have been detected in in vitro–produced porcine embryos, whereas in vivo–produced embryos and embryos produced by nuclear transfer showed higher expression levels (Kumar et al.,2007). However, in the same study, NANOG mRNA levels in in vivo–derived blastocysts were approximately equal to in vivo–derived 4-cell-stage embryos. Moreover, differences in mRNA levels were less than 2-fold between all in vivo–derived stages, which included 4-cell-stage embryos, 8-cell-stage embryos, morulae, and blastocysts. These observations, in combination with those from the present study, do not support a large role for NANOG in early lineage segregation events or in defining the pluripotent cell population in pig embryos. On the other hand, the heterogeneous distribution observed in ICMs of bovine blastocysts, which resembles that of mouse (Chazaud et al.,2006), suggests a conserved function of this gene between these species.

From germinal vesicle to metaphase 2-stage oocytes, OCT4 showed upregulation, but at the 2-cell stage, mRNA levels were reduced again below those of the germinal vesicle stage. Oct4 is one of the four factors that can reprogram somatic nuclei (Takahashi and Yamanaka,2006). Moreover, the peak observed in its expression profile coincides with the stage at which oocytes can reprogram a somatic nucleus (Gao et al.,2002). Possibly high levels of OCT4 in metaphase 2-stage oocytes indicate a role for OCT4 in oocyte maturation, preparing the oocyte for totipotency. An alternative explanation could be the building up of maternal mRNA stores. The potential role of OCT4 at these early stages can be studied by interfering with OCT4 expression in oocytes. Does this affect the ability of oocytes to reprogram somatic nuclei and what is the effect on the epigenetic status of the maternal genome?

Porcine blastocysts showed increased mRNA expression of OCT4 and, in line with previous findings (Kirchhof et al.,2000), in vitro– as well as in vivo–produced blastocysts expressed OCT4 protein in nuclei of both the ICM and the TE. Remarkably, and in contrast with mouse embryos (Palmieri et al.,1994; Niwa et al.,2005), in vivo–produced porcine morulae lacked such nuclear OCT4 expression, which is another indication that this protein is not involved in inhibition of TE formation as it is in the mouse (Niwa et al.,2000). Proliferation of TE stem cells depends on fibroblast growth factor 4 (FGF4) (Tanaka et al.,1998), the expression of which is under the control of a complex formed by OCT4 and SOX2, a member of the Sry-related Sox factor family (Yuan et al.,1995). It has been suggested that for species with an epitheliochorial placenta, continued OCT4 expression in TE cells is essential to stimulate FGF4-mediated self-renewal of TS stem cells, which allows elongation and prevents premature differentiation of the trophectoderm (Degrelle et al.,2005; He et al.,2006). This idea is supported by the finding that OCT4 expression is not restricted to the ICM in caprine embryos (He et al.,2006).

CDX2, which is involved in formation of the mouse TE (Niwa et al.,2005; Strumpf et al.,2005), was expressed at higher levels in porcine blastocysts compared to earlier developmental stages, and congruous with this CDX2 was expressed in TE cells of pig blastocysts as well as bovine blastocysts. Therefore, CDX2 is likely to be involved in the formation of the TE in these species.

Surprisingly, expression of GATA4, important for the formation of the PE in mouse, was significantly reduced in pig blastocysts compared with earlier developmental stages, which suggests that it is not involved in major processes that occur in blastocysts. Indeed, GATA4 protein was not detected in porcine blastocysts. GATA6 on the other hand, was expressed at significant higher levels in blastocysts. In porcine in vivo–produced blastocysts and in bovine blastocysts, GATA6 was localized to a subset of cells of the ICM. As a consequence, GATA6 is expected to play a role in PE formation in porcine and bovine embryos similar to its function in mouse embryos (Chazaud et al.,2006). Absence of GATA6 in in vitro–produced porcine embryos could indicate differences in embryos as a result of their origin, but could also be a reflection of differences in developmental age.

UTF1 and CK18 are specifically expressed in blastocyst stages (Brulet et al.,1980; Oshima et al.,1983; Okuda et al.,1998). These genes served as positive controls for the QPCR and conform the expectations, UTF1 and CK18 showed significant increased expression in blastocysts.

In summary, OCT4 and NANOG behave differently in pig embryos (Fig. 5) than in mouse embryos, where these factors play a role in the formation of the pluripotent primitive ectoderm. This makes it unlikely that OCT4 and NANOG are involved in the specification of the primitive ectoderm or defining the pluripotent cell population. In bovine embryos, however, the protein NANOG showed a similar random-like distribution in the cells of the ICM as in mouse embryos, which indicates a role for NANOG in the development of the PE in bovine embryos. CDX2 expression in porcine and bovine embryos resembled that of mouse embryos, suggesting a conserved role for CDX2 in the formation of the TE between mammals. A random distribution of GATA6 in porcine and bovine cells of the ICM, comparable to that found in mouse, also implies a conserved function of GATA6 in PE formation in mammalian development. From these experiments, it can be concluded that mammals differ in early lineage segregation, which potentially influences the formation and characteristics of the pluripotent cell population. Functional studies such as siRNA-mediated knockdown are needed to further elucidate the roles of factors such as OCT4, NANOG, CDX2, GATA4, and GATA6 in other species than the mouse.

Figure 5.

Early lineage segregation in mouse, pig, and cattle. See text for details.


Oocyte Maturation, IVF, Embryo Culture, and Collection of In Vivo–Produced Embryos

All incubations described below took place in a humidified atmosphere of 38.5°C and 5% CO2, unless noted otherwise. Oocyte retrieval, in vitro oocyte maturation, in vitro fertilization, and subsequent in vitro culture of porcine embryos proceeded as previously described (Kidson et al.,2003). In short, cumulus oocyte complexes (COCs) were aspirated from 2–6-mm antral follicles of sow ovaries, which were collected at a regional slaughterhouse. Equally sized oocytes with at least three layers of compact cumulus were selected and placed in maturation medium, containing NCSU-23 medium (Petters and Reed,1991), supplemented with 10% porcine follicular fluid, 0.57 mM cysteine, 25 mM β-mercaptoethanol, 10 IU/ml eCG (Chorulon, Intervet, Boxmeer, The Netherlands), and 10 IU/ml hCG (Folligonan, Intervet). After 24-hr culture, the COCs were transferred to hormone free maturation medium and cultured for an additional 18 hr. After culture, denuded oocytes were placed in Tris-buffered IVF-medium, containing 113.1 mM NaCl, 3 mM KCl, 20 mM Tris, 11 mM D-glucose, 1 mM caffeine, 5 mM sodium pyruvate, 7.5 mM CaCl2, 0.1% (w/v) bovine serum albumine (BSA; Sigma-Aldrich Chemie, Zwijndrecht, The Netherlands), and 1% pen/strep. Oocytes were fertilized in vitro with fresh semen from two randomly selected boars at a concentration of 1,000 cells/oocyte. From one day after IVF onward, presumptive zygotes were cultured in IVC medium, which contained NCSU-23 and 0.4% BSA.

An estrous sow was artificially inseminated 5 days after weaning and sacrificed at day 5 of pregnancy. Subsequently, the genital tract was removed and flushed with PBS to collect in vivo–derived embryos. The Institutional Animal Care and Use Committee of Utrecht University approved of this animal experiment.

Bovine embryos were retrieved from in vitro cultures, using the following methods. COCs were retrieved from 3- to 8-mm follicles of ovaries that were supplied by a local slaughterhouse. Groups of 35 COCs with an intact cumulus oophorus were incubated in 500 μl M199 (Gibco BRL, Paisley, UK) supplemented with 10% fetal calf serum (Gibco BRL), 0.2 IU/ml bovine FSH (Sioux Biochemical Inc., Sioux Center, IA), 0.2 IU/ml bovine LH (Sioux Biochemical Inc.), 15.42 μg/ml cysteamine (Sigma-Aldrich), and 1% (v/v) penicillin-streptomycin (Gibco-BRL). In vitro fertilization was performed 23 hr after oocyte maturation according to the procedure previously described by Parrish et al. (1988) with a few modifications (Izadyar et al.,1996). In short, oocytes were transferred to fertilization medium (Fert-Talp) supplemented with 1.8 IU/ml heparin (Sigma-Aldrich), 20 μM d-penicillamine (Sigma-Aldrich), 10 μM hypotaurine (Sigma-Aldrich), and 1 μM epinephrine (Sigma-Aldrich). Frozen-thawed semen from a fertile bull was centrifuged over a Percoll gradient and sperm cells were added to a final concentration of 5 × 105 spermatozoa/ml. After 20 hr of incubation, presumptive zygotes were denuded by vortexing for 3 min and subsequently placed in synthetic oviduct fluid (SOF) medium. Incubation took place at 39°C in a humidified atmosphere with 7% O2 and 5% CO2. On day 5 of embryo culture, cleaved embryos were transferred to fresh SOF medium and blastocysts were collected on day 8.

RNA Extraction and Reverse Transcription

Total RNA was isolated from denuded germinal vesicle stage oocytes, denuded metaphase 2-stage oocytes (as confirmed by the presence of one polar body), 2-cell-stage embryos, 4-cell-stage embryos, early blastocysts, and expanded blastocysts, using the RNeasy minikit (Qiagen, Venlo, The Netherlands). RNA yield is higher in blastocyst stage embryos compared to oocytes and cleavage stage embryos and, therefore, 40 oocytes/embryos per sample were collected for the early stages and 10 blastocysts were pooled for every sample. Primed with random primers, RNA was reverse transcribed to first-strand cDNA with Superscript II (Invitrogen, Groningen, The Netherlands). For each sample, mixtures were also prepared without reverse transcriptase to serve as negative controls. Synthesis of cDNA was carried out for 1 hr at 42°C, after which samples were placed at 80°C for 5 min, chilled on ice, and stored at −20°C.

Quantitative RT-PCR

Prior to quantitative reverse transcription-polymerase chain reaction (QPCR) amplification, primers were designed using Primer Select software (DNAstar, Madison, WI) and Beacon Designer 4 (PREMIER Biosoft International, Palo Alto, CA) (Table 1). Primers were tested on genomic DNA, cDNA of porcine oocytes, and in vitro–produced embryos. The amplicons were run on 2% agarose gels and primer specificity was confirmed by sequencing of the products. Subsequent PCR was executed in a Bio-Rad iCycler (Bio-Rad, Veenendaal, The Netherlands). The reaction and quantification of the transcripts was performed with iQ SYBR Green supermix (Bio-Rad). With a temperature gradient, the optimal annealing temperature of each primer pair was determined. Every gene was run separately and a standard curve of tenfold dilutions ranging from 10 pg to 1 ag supplemented each run. Three technical replicates were run of all points of the standard curve and all samples, and 1 μl/reaction of cDNA was used as template. This sample was added to 24 μl of reaction mixture, containing 12.5 μl H2O, 11.25 μl iQ SYBR Green supermix (Bio-Rad), and 0.5 mM of both forward and reverse primers (Isogen, Maarssen, The Netherlands). The thermal cycling profile started with a 3-min dwell temperature of 94°C, which was followed by 40 cycles with 4 steps/cycle; 30 sec at 94°C, 30 sec at the primer-specific annealing temperature, 30 sec at 72°C, and finally a step at which fluorescence was acquired. These cycles were followed by a post-dwell of 1 min at 94°C and, finally, a melt curve was generated by temperature increments of 0.5°C starting from 65 to 100°C, with fluorescence acquisition after each step.

Table 1. Primers Used for RT-PCR and Q-PCRa
GeneGenBank numberPrimersTa (°C)Amplicon size (bp)
  • a

    OCT4 primers were based on primers specific for bovine OCT4, which has been previously described by Van Eijk et al. (1999). The sequence of this gene was blasted against porcine nucleotide sequences, which resulted in one sequence with high query coverage. After alignment of this porcine with the bovine sequence, bovine specific primers were redesigned by nucleotide substitution of any mismatching nucleotides. UTF1 primers were designed on a pig sequence that has 91% coverage with the coding sequence of Homo sapiens mRNA for UTF1. CDX2 was detected with E-PCR primers from the UniSTS database on the NCBI website (UniSTS code: RH48331). GATA6 primers have been described by Gillio-Meina et al. (2003).

OCT4AJ251914Forward: 5′-GTTCTCTTTGGGAAGGTGTT-3′55.4313
UTF1CN028152Forward: 5′-CCGCGGGCCCGACCTCACG-3′66.0216
CDX2EU137688Forward: 5′-GTCACCAGAGCTTCTCTGGG-3′52.9144
CK18EU131884Forward: 5′-ATGAAGAAGAACCACGAGGAGGAA-3′54.8118
GATA4NM_214293Forward: 5′-ATGAAGCTCCATGGTGTCCC-3′55.8162

Data were analyzed with IQ5 software (Bio-Rad), with which the starting quantities of all candidate genes were calculated, based on their standard curves. QPCR data was normalized to GAPDH, PGK1, S18, and UBC, which have been demonstrated as a good set of reference genes for QPCR studies in porcine oocytes and preimplantation embryos (Kuijk et al.,2007).

Dilution curves of all candidate reference genes showed an average amplification efficiency of 93.6% ± 10.3 and an average coefficient of determination (R2) of 0.986 ± 0.013. Single distinctive peaks in the melt curves verified specific amplification of the gene of interest. Integrity of the cDNA samples was confirmed by consistent detection of all reference genes in all samples, except for one metaphase-2 sample, which was excluded from further analysis. Genomic DNA contributions, as determined by −RT levels, were 0 in the case of OCT4, and detected at an average level of 0.44% for CDX2, 0.63% for CK18, 0.012% for GATA4, 0.033% for GATA6, and 0.64% for UTF1. If −RT levels were higher than 5% of the +RT levels, samples were excluded from the analysis for that particular gene, which resulted in exclusion of 5 samples for CDX2, 3 samples for CK18, and 3 samples for GATA6, without loss of a developmental stage for any of these genes.

For each developmental stage, normalized values were divided by the normalized value of the germinal vesicle stage for that particular gene. As a consequence, gene expression levels are relative to the amount of expression in oocytes. Values for blastocyst stages were compared with the mean expression level and corrected standard deviation of all earlier stages. For OCT4, blastocyst stages were compared with cleavage stages. Statistical differences between both groups were tested by Student's t-test and a probability value <0.05 was considered significant. For OCT4 and GATA4, a Bonferroni correction was applied for multiple comparisons and for these genes a probability value of <0.025 was considered significant.

Whole Mount Immunofluorescence

In vitro– and in vivo–derived embryos were fixed overnight with 4% paraformaldehyde in PBS with 0.2% Tween (ICN Biomedicals). Subsequently, embryos were permeabilised in methanol at −20°C overnight and Tris buffered saline with 0.05% Tween (TBST) and 0.1% TritonX (Sigma-Aldrich) for 10 min, after which the embryos were blocked for 1 hr in TBST with 0.5% BSA. Next, the embryos were incubated in primary antibody 3–4 days, followed by another 2 days of incubation in the secondary antibody. Antibodies (Table 2) were diluted in blocking solution and rabbit and mouse isotypes served as negative control. Embryos were counterstained with ToPro (Invitrogen) and mounted in Vectashield (Vectorlab, Burlingame, CA) before observation. Fluorescent signals were visualized using a Confocal Laser Scanning Microscope (Bio-Rad) and on an Olympus BH2 epifluorescence microscope.

Table 2. Antibodies Used for Immunofluorescence
ImmunogenaSourceConcentration (μg/ml)Description
  • a

    AA, amino acids.

Full length mouse Nanog fusion proteinAB21603 (Abcam)2Rabbit polyclonal
AA 1-134 of human Oct4SC5279 (Santa Cruz)4Mouse monoclonal
Synthetic peptide corresponding to AA 234–248 of human Cdx2AB4123 (Chemicon)2Rabbit polyclonal
C-terminal AA 328–439 of human GATA-4SC9053 (Santa Cruz)4Rabbit polyclonal
C-terminal AA 358-449 of human GATA-6SC9055 (Santa Cruz)4Rabbit polyclonal
Rabbit IgG controlSC2027 (Santa Cruz)4Rabbit
Mouse IgG controlSC2025 (Santa Cruz)4Mouse


We would like to thank Anko M. de Graaff and Richard W. Wubbolts of the Center for Cell Imaging at the Faculty of Veterinary Medicine in Utrecht for assistance in confocal laser scanning microscopy. We would also like to thank Susana M. Chuva de Sousa Lopes of the Hubrecht Institute in Utrecht for her assistance in the whole mount immunofluoresence procedure.