Introduction
With the rapid development of life sciences and the biological medicine industry, the need for selecting suitable laboratory animals has become more urgent. The domestic pig is an important livestock species and its lifespan and physiology are far more comparable with those of humans than are those of rodents. Therefore, it is a worthwhile animal modal for biomedical and developmental studies.
Early embryonic development is the fundamental process for further complex biological events. Patten (Reference Patten1948) described the embryonic development in pig, and his excellent work can still provide some fundamental knowledge for current research. Vejlsted et al. (Reference Vejlsted, Offenberg, Thorup and Maddox-Hyttel2006a,b) employed a stereomicroscopic staging system to clearly define the six stages of porcine embryonic development from hatching to preimplantation (days 8–17), and provided morphological details for every stage. Hassoun et al. (Reference Hassoun, Schwartz, Feistel, Blum and Viebahn2009) defined three typical stages during early gastrulation of porcine embryos by stepwise morphological analysis of two axial structures and also used in situ hybridization to detect three key genes involved in early embryonic development. Although these studies provide comparable information for early morphological development of porcine embryos, it is still not enough for current requirements, and far less than that available for the mouse. In the pig, the first lineage segregation occurs during the morula to blastocyst stage, which is around day 5 of in vivo embryos, and it is easy to observe the differentiated inner cell mass (ICM) and trophoblast (TE) under a stereomicroscope (Perry & Rowlands, Reference Perry and Rowlands1962). At day 7, most blastocysts have escaped from the zona pellucida (Barends et al., Reference Barends, Stroband, Taverne, Kronnie, Leën and Blommers1989). As the blastocyst hatches, it becomes impossible to define all differentiated cell groups by observation under a stereomicroscope. At days 11–13, the embryos have formed the primitive streak and started to migrate (Vejlsted et al., Reference Vejlsted, Offenberg, Thorup and Maddox-Hyttel2006b). So, this time period (days 7–13) is the period of embryonic disc formation and mesoderm differentiation. In addition, the porcine structure of the embryonic disc is different from that in the mouse, and it is challenging to trace this process in detail. Hence, focusing on dynamic morphological changes combined with molecular evidence of specific embryonic stages will provide more information to further understand the differentiation of porcine germ layers.
At this time, the associated regulating mechanisms of pigs have gradually drawn particular attention. However, many questions still remain over porcine embryonic differentiation. GATA6 is used as a molecular marker for the hypoblast in mice, cattle, monkeys and humans, and also as a marker that identifies the parietal endoderm in mouse embryos (Koutsourakis et al., Reference Koutsourakis, Langeveld, Patient, Beddington and Grosveld1999; Kuijk et al., Reference Kuijk, Du Puy, Van Tol, Oei, Haagsman, Colenbrander and Roelen2008). Although studies of GATA6 expression in porcine blastocysts have demonstrated that it resembles that in mouse and bovine blastocysts, having a random ‘salt and pepper’ pattern (Kuijk et al., Reference Kuijk, Du Puy, Van Tol, Oei, Haagsman, Colenbrander and Roelen2008; Kumar et al., Reference Kumar, Maeng, Jeon, Lee, Lee, Jeon, Ock and Rho2012; Rodríguez et al., Reference Rodríguez, Allegrucci and Alberio2012), the critical specific expression of GATA6 in extra-embryonic endoderm of gastrulation still cannot be detected, so it is necessary to detect the dynamic expression of GATA6 following differentiation.
OCT4 (Pou5f1), a typical pluripotency marker of the Pit-Oct-Unc family, plays important key roles in first lineage segregation (ICM/TE), and is used as a marker for pluripotent cells in mouse development both in vivo and in vitro (Ginis et al., Reference Ginis, Luo, Miura, Thies, Brandenberger, Gerecht-Nir, Amit, Hoke, Carpenter, Itskovitz-Eldor and Rao2004). Also, SOX2 is essential to maintain the self-renewal of pluripotent ICM and embryonic stem (ES) cells in mouse (Avilion et al., Reference Avilion, Nicolis, Pevny, Perez, Vivian and Lovell-Badge2003). In pig, although some researchers have found controversial OCT4 expression patterns in both ICM and TE cells of early porcine blastocysts (Kirchhof et al., Reference Kirchhof, Carnwath, Lemme, Anastassiadis, Schöler and Niemann2000; Vejlsted et al., Reference Vejlsted, Du, Vajta and Maddox-Hyttel2006a; Kuijk et al., Reference Kuijk, Du Puy, Van Tol, Oei, Haagsman, Colenbrander and Roelen2008; du Puy et al., Reference du Puy, Lopes, Haagsman and Roelen2011; Wolf et al., Reference Wolf, Rasmussen, Schauser, Jensen, Schmidt and Hyttel2011a, Reference Wolf, Serup and Hyttelb), for later development, OCT4 expression is lost in the TE and only expressed in the nuclei of epiblast cells, followed by expression in the intra- and extra-embryonic mesoderm (du Puy et al., Reference du Puy, Lopes, Haagsman and Roelen2011; Wolf et al., Reference Wolf, Serup and Hyttel2011b). For SOX2, du Puy et al. (Reference du Puy, Lopes, Haagsman and Roelen2011) reported that its mRNA and protein are not expressed in the ICM of the early blastocyst at around day 6, but that the protein begins to be expressed in ICM at day 8.5. Both mRNA and the protein are detectable in the epiblast of embryos at days 10 and 11 for pig (Hall et al., Reference Hall, Jacobsen, Rasmussen and Hyttel2010; du Puy et al., Reference du Puy, Lopes, Haagsman and Roelen2011; Gao et al., Reference Gao, Hyttel and Hall2011). In general, these data indicate that OCT4 and SOX2 have a different expression characteristic in porcine embryos compared with that in mouse embryos. Further studies are required to confirm this early differentiation of pig.
For mesoderm differentiation, VIMENTIN is often used as a marker for mesenchymal-derived cells and cells that are undergoing epithelial-to-mesenchymal transition in both normal development and metastatic progression. VIMENTIN is a type III intermediate filament protein that is expressed in mesenchymal cells, and is the major component of their cytoskeleton (Yamada et al., Reference Yamada, Kawamata, Walker and McGeer1992). In porcine embryos, some data indicate that VIMENTIN is expressed in the ICM of the early blastocyst, and then its expression is lost when the ICM differentiates into the hypoblast and epiblast (Hall et al., Reference Hall, Jacobsen, Rasmussen and Hyttel2010). However, at around day 8, VIMENTIN reappears in the epiblast (Hall et al., Reference Hall, Jacobsen, Rasmussen and Hyttel2010) and, together with gastrulation, it is expressed in the mesenchyme derived from the epiblast, which extends between the embryonic disc and underlying visceral endoderm (Fléchon et al., Reference Fléchon, Degrouard and Fléchon2004). The dynamic expression pattern of VIMENTIN in porcine embryos indicates a conserved role in mesoderm differentiation between the mouse and pig. Therefore, it can be used to trace the differentiation model of mesoderm in pig.
Therefore, the purpose of the current study was to gain an insight into the development of in vivo porcine preimplantation embryos (days 7–13) with special focus on the differentiation of the hypoblast and mesoderm by semi-thin section. In addition, molecular markers GATA6, OCT4, SOX2 and VIMENTIN were detected by whole mount immunohistochemistry to understand further the differentiation processes of porcine embryos. The current study may provide a better understanding of early embryonic development in pig.
Materials and methods
Embryo collection
In total, 171 embryos were collected from 14 Yorkshire-Danish Landrace sows from days 6.5–13 post-insemination. The sows were mated twice within 24 h after detection of oestrus, at an interval of 12 h. The last mating time was taken to be day 0 (D0) of conception. Reproductive organs collected from the sows were transported to the laboratory in warm saline (37°C) for a maximum period of 1 h. The embryos were rinsed twice using 50 ml phosphate-buffered saline (PBS) that contained 5% fetal bovine serum (FBS). Whole embryos or the region of the embryonic disc after days 11 to 13 (D11–D13) were photographed under a stereomicroscope, staged and subsequently fixed for semi-thin sectioning and immunohistochemistry. All procedures were approved by the Animal Ethics Committee of the Northeast Agriculture University, Harbin, China.
Processing of embryos for semi-thin sections
Blastocysts or the embryo proper were fixed in 0.1 M Na-phosphate buffer (pH 7.4) that contained 3% glutaraldehyde (Sigma-Aldrich) for at least 2 h at 4°C, post-fixed in 1% osmium tetroxide (Sigma-Aldrich), and dehydrated in ascending concentrations of alcohol (50–100%). Then, embryonic tissues were displaced by pyruvate and individually embedded in Epon 812. Complete series of semi-thin sections were cut at 1-μm intervals using a diamond knife and then mounted on slides. To remove the epoxy resin, the semi-thin sections were treated in a saturated solution of sodium hydroxide for 5 min at room temperature. Then, the sections were successively dehydrated by a gradient alcohol series and rinsed 3–4 times in distilled water. The sections were stained with 1% toluidine blue, rinsed in distilled water, air dried, submersed in xylene, and coverslipped using a xylene-based neutral balsam.
Whole mount immunofluorescence
Embryos were fixed with 4% paraformaldehyde in PBS for 40 min at room temperature. Samples were permeabilized with 1% Triton X-100 (Sigma-Aldrich) in PBS for 5 h at 4°C, and then blocked with 1% bovine serum albumin (BSA) in PBS at room temperature before incubation with the indicated antibodies. Next, samples were incubated with primary antibodies in 0.01% Triton X-100 and 0.1% Tween 20 in PBS overnight at 4°C. The antibodies used in the current study were polyclonal goat anti-OCT4 (1:100, SC8628; Santa Cruz Biotechnology), polyclonal goat anti-SOX2 (1:100, SC17320; Santa Cruz Biotechnology), and polyclonal rabbit-GATA6 (1:100, Ab22600; Abcam). Embryos were washed at least three times for 20 min with 0.01% Triton X-100 and 0.1% Tween 20 in PBS at room temperature, and then incubated with secondary antibodies diluted at 1:1000 in 0.01% Triton X-100 and 0.1% Tween 20 in PBS for 1 h at room temperature. The secondary antibodies used were donkey anti-rabbit (A10040; Invitrogen) and donkey anti-goat (A11055; Invitrogen). Finally, embryos were counterstained with Hoechst 33342 (Sigma-Aldrich), mounted on glass slides, and examined under a Nikon microscope and laser-scanning confocal microscope.
Whole mount immunohistochemistry
Whole mount immunohistochemistry was performed according to Nagy et al. (Reference Nagy, Gertsenstein, Vintersten and Behringer2002). Collected embryos were fixed in freshly prepared solution of methanol and dimethyl sulphoxide (DMSO; 4:1) overnight at 4°C (>10 h). Subsequently, the samples were permeabilized with methanol–DMSO–H2O2 for 5–10 h at 4°C to inactivate endogenous peroxidases, and then dehydrated in 100% methanol at –20°C for several months. Embryos were rehydrated and shaken for 30 min in 50% methanol–PBS, 30 min in PBS, and then 30 min in PBS–skimmed milk–Tween (PBSMT) twice before incubation overnight at 4°C with primary antibodies, anti-OCT4 (SC-8628; Santa Cruz Biotechnology) or anti-VIMENTIN (V6630; Sigma-Aldrich), diluted at 1:100 in PBSMT. The next day, samples were washed three times in PBSMT for 1 h and then incubated with the secondary antibodies, rabbit anti-goat (Ab6741; Abcam) or goat anti-mouse (A9044; Sigma-Aldrich), diluted at 1:300 in PBSMT for about 10 h at 4°C. Finally, embryos were stained with DAB (Ab94665, Abcam) until development of an appropriate signal, washed twice in PBS for 5 min each, and then post-fixed in 3% glutaraldehyde for whole embryos and semi-thin sections.
Results
General morphological changes upon development of hatching porcine embryos
The sows were mated twice within 24 h after detection of oestrus. The last mating time was taken to be D0 of conception. At around D6.5, blastocysts were surrounded by the zona pellucida (ZP), in which the ICM was visible and surrounded by the TE at one pole (Fig. 1A). From D7 onwards, embryos escaped from the ZP and were spherical with a diameter of 187.5—250.0 μm. The embryo proper was completely covered by the polar trophoblast (Rauber's layer), which was hard to identify in the living embryos (Fig. 1B). The embryo proper was not visible at about D8 under a light microscope. Therefore, to verify the embryo proper, blastocysts were fixed in osmic acid. In the area opaca, it was found that the embryo proper was covered by a darker Rauber's layer (Fig. 1C). Next, it was found that the morphology of the conceptus varied between spherical (Fig. 1C, E, F), ovoid (Fig. 1D) and tubular (Fig. 1G, H) at D8–D11 as observed by stereomicroscopy. With the disappearance of Rauber's layer in these embryos, either partly (Fig. 1E) or completely (Fig. 1F, G), a translucent embryo proper area was observed. Conceptus development was characterized by rapid trophoblastic elongation at D11–D12. At around D11, some conceptuses were filamentous and the embryo proper was distinctively different from the extra-embryonic region in the dorsal view. In addition, the anterior--posterior axis emerged in the embryo proper. At D12, the most conceptuses were filamentous and longer than 17 cm, the embryo proper was convex and located in the midline of the conceptus (Fig. 1J), and the primitive streak emerged in the centre of the embryo proper along the anterior–posterior axis (Fig. 1J′). Around D13, the conceptus was elongated to over 74 cm and the embryo proper was present in the conceptus in the shape of a broad-brimmed rain hat (Fig. 2G) and a pit apparently existed in the midline of the primitive streak (Figs. 1K and 2G).
Morphological changes upon development of hatching porcine embryos. Sows were mated twice within 24 h after detection of estrus. The last mating time was taken to be day 0 (D0) of conception. At about D6.5 (A), blastocysts were surrounded by the zona pellucida (ZP), and the inner cell mass (ICM) was visible (arrow). At D7 (B), all embryos escaped from the ZP, and were spherical. Afterward, the morphology of the conceptus varied between spherical (B, C, E, F), ovoid (D), tubular (G, H), and filamentous (J) during D7–D13 as shown by stereomicroscopy. With partial or complete disappearance of Rauber's layer (E, arrowhead), a translucent area, namely the embryo proper, was identified at D8.5–D11 (arrow). At D7–D8.5 (B–D), the embryo proper was covered by Rauber's layer, and hard to discern (B, C). When the embryo was fixed in osmic acid (C), the darkfield (embryo proper, box) was confirmed. At about D11–D12 (G, H), some embryos were tubular, which transiently converted to filamentous. At about D11–D12 (I, J), the majority of conceptuses were filamentous and longer than 17 cm, and the embryo proper was convex, located in the midline of the conceptus (J, box), and had an anterior–posterior axis (I). The primitive streak was observed at D12–D13 (J′, K, hollow arrowhead). At about D13 (K), the primitive streak was depressed in the centre (hollow arrowhead).
Morphological character of embryo proper development in preimplantation embryos. (A) Formation of parietal endoderm. The darkfield (A) (box) was the embryo proper area with an obscure boundary. Histological sections revealed that that the PE migrated along the TE and formed a continuous, but fenestrated, layer lining the blastocoele (a2), and visceral endoderm showed anterior visceral endoderm (a1). (B) A cavity appeared in the epiblast with Rauber's layer. The oval outline of the embryo proper could be defined in the dorsal view after osmic acid fixation, although it was still overlaid by Rauber's layer. Serial sections showed that a space appeared in the centre of the epiblast, and Rauber's layer was still present (b1, b2). (C) The embryonic disc began to establish, which was accompanied by partial loss of Rauber's layer, and a translucent embryo proper area was identified in the dorsal view. Serial sections (c1, c2) showed that the embryonic disc began to establish with formation of the pre-amnion (decagon). (D) Shape of the embryonic disc. The embryonic disc was established with disappearance of Rauber's layer after fixation by osmic acid in the dorsal view (D; around D9). In addition, the epiblast began to thicken and form the primitive streak (d1; d2). (E) Mesoderm differentiation started in the embryonic disc. A crescent-shaped thickening of the caudal portion of the embryonic disc appeared in the dorsal view (arrow). Sections of the same embryo confirmed an increased thickness of the epiblast and cells appeared loosely at the posterior of the primitive streak (e). (F) Mesoderm extension under the epiblast. The embryo proper was pear-shaped and convex in the posterior view. Mesoderm cells migrated from the epiblast through the basal membrane and extended toward the extra-embryonic area (f). (G) A chord emerged under the centre of the epiblast, and a pit apparently existed in the midline of the primitive streak (arrowhead). As the chord appeared in the centre, lateral mesoderm cells migrated along the trophoblast and PE, and the extra-embryonic coelom was formed (g). The position and orientation of the sections are indicated with bars. Asterisks are placed just peripheral to the embryonic disc border. Dots delineate the area of the anterior hypoblast with the high columnar and dense region. P to A: posterior–anterior axis. aVE: anterior visceral endoderm; Ch: chord; Epi: epiblast; Exe: extra-embryonic coelom; Me: mesoderm; pEn: parietal endoderm; PS: primitive streak; Rb: Rauber's layer; TE: trophoblast; VE: visceral endoderm.
Structural character of embryo proper development in preimplantation embryos
To describe the morphology and structural character of hatching embryos, all samples (Fig. 1B–F) were fixed in 3% glutaraldehyde, re-fixed in osmic acid, and then prepared as semi-thin sections. The ICM cells with uniform cytoplasm were closely aligned in newly hatched embryos. The hypoblast was simple with flattened discernible cells underlying the ICM in histological sections, and the embryo proper was surrounded by TE cells with a high content of intracytoplasmic coarse granules (data not shown). Upon formation of the embryonic disc, developmental processes were identified in the sections as follows:
1. Formation of parietal endoderm. The darkfield embryo proper area had an obscure boundary, and was covered by a darker Rauber's layer in the dorsal view after osmic acid fixation (Fig. 2A). Histological sections revealed that the hypoblast, namely primitive endoderm (PE), migrated along the TE and formed a continuous, but fenestrated, layer that lined the blastocoele, where hypoblast cells were flat and elongated, and defined as parietal endoderm (Fig. 2a2). Visceral endoderm underlying the epiblast showed closely packed simple epithelial cells and differentiation had started anteriorly (Fig. 2a1). Epiblast cells had homogeneous cytoplasm and a non-polar character, which were covered by Rauber's layer with cells containing coarse intracytoplasmic granules;
2. A cavity appeared in the epiblast that was covered with Rauber's layer. The oval outline of the embryo proper (darkfield) could be defined in the dorsal view after osmic acid fixation, although it was still overlaid by Rauber's layer (Fig. 2B). In two out of six embryos, serial sections showed that an irregular space appeared in the centre of the epiblast, and a thin coarse Rauber's layer still existed (Fig. 2b1, b2).
3. The embryonic disc began to establish, accompanied by partial loss of Rauber's layer, and a translucent embryo proper area could be identified in the dorsal view (Fig. 2C). In one out of three embryos, serial sections showed that the epiblast became thick and formed the embryonic disc, and the cavity partially disappeared with loss of Rauber's layer. The cells of the embryonic disc were columnar epithelial cells. Moreover, anterior visceral endoderm cells still existed in the visceral endoderm (Fig. 2c1). In the same blastocyst, the disappearing cavity was still visible, surrounded by epiblast cells, and overlaid by some degenerating Rauber's layer cells (Fig. 2c2).
4. Shape of the embryonic disc. A light embryo proper region was recognized under the microscope (Fig. 1F). The embryo proper emerged as a long oval shape, and the darker Rauber's layer disappeared after fixation with osmic acid in the dorsal view (Fig. 2D). Histological sections indicated that the embryonic disc consisted of the epiblast and hypoblast. In addition, the epiblast began to thicken and formed the primitive streak (Fig. 2d1–2).
In general, the embryonic disc was established through the four processes described above. During this period, the hypoblast displayed anterior visceral endoderm. Also, a space appeared transiently among the epiblast cells with the disappearance of the cavity and formation of the embryonic disc.
Next, the embryonic disc started to form the primitive streak and mesoderm proceeding in gastrulation. In the current study, a crescent-shaped thickening of the caudal portion of the embryonic disc appeared in the dorsal view after osmic acid fixation (Fig. 2E). Sections of the same embryo confirmed the increased thickness of the epiblast that was lined with a basal membrane. Asymmetric differentiation with a thickened posterior was observed in the epiblast, and irregular loosely packed cells appeared at the caudal portion of the primitive streak (Fig. 2e). Subsequently, the mesoderm extended under the epiblast. The embryo proper was pear-shaped and convex in the posterior view (Fig. 2F). Mesoderm cells differentiated from the epiblast through the basal membrane and extended toward the extra-embryonic area (Fig. 2f). At around D13, a cord emerged under the centre of the epiblast. A pit apparently existed in the midline of the primitive streak (Fig. 2G). As the cord appeared in the centre, lateral mesoderm cells migrated along the trophoblast and PE to form the parietal and visceral mesoderm, respectively, which were verified in transverse sections. Finally, the extra-embryonic coelom was formed, and the epiblast and PE were displaced by the ectoderm and definitive endoderm, respectively (Fig. 2g).
Localization of GATA6, OCT4 and SOX2 in secondary lineage segregation
To confirm the differentiation characteristics, immunohistochemical staining of the endoderm marker GATA6 was performed according to the outcomes of mouse embryonic development. As a result, GATA6 expression was hardly visible in newly hatched embryos. In parallel, real-time polymerase chain reaction (RT-PCR) was used to confirm the results (data not shown). Next, transcription factors OCT4 and SOX2, essential for the maintenance of ES cells and the ICM, were investigated. OCT4 was stained in the ICM and most trophoblast cells (Fig. 3A) in newly hatched embryos, but SOX2 was only observed in the ICM (Fig. 3B).
Germ layer-specific gene expression in endoderm lineage segregation of hatched embryos. (A, B) In newly hatched embryos, OCT4 was expressed in the ICM and TE, but SOX2 was only expressed in the ICM. (C–E) GATA6 expression in the extra-embryonic area indicated parietal endoderm formation. The embryonic proper was verified by OCT4 or SOX2 expression. (F, G) OCT4 expression concentrated in the forming embryonic disc, and the polar TE (Rauber's layer) lost expression of OCT4. The newly forming primitive endoderm (hypoblast) also had no OCT4 expression. Serial sections revealed that Rauber's layer was disappearing (f1, f2) and the primitive disc (g1, g2) was forming. The region between the asterisks indicates the embryonic disc border (g1, g2). The embryos were co-stained for OCT4 and GATA6 in (C) and (D), and the embryo in (E) was co-stained for SOX2 and GATA6. Dotted lines indicate the embryonic proper, and the upper right in (C) is the whole embryo. DNA was labelled with Hoechst (HO). Epi: epiblast; TE: trophoblast; VE: visceral endoderm.
At the stage of parietal endoderm formation, the embryonic disc was establishing and Rauber's layer was disappearing. GATA6 positivity was found in the extra-embryonic area that had small squamous cells with hyperchromatic nuclei (Fig. 3C–E), but it was negative in the embryo proper and larger cells with hypochromatic nuclei. These results indicated that GATA6 was expressed in parietal endoderm at about D8–D9. At this time, both OCT4 and SOX2 were expressed in the embryo proper, except for the extra-embryonic area and trophoblast cells at about D8–D9 (Fig. 3D, E). Before embryonic disc establishment, whole mount immunohistochemistry and embryo sections were used to verify OCT4 localization in the embryo proper covered by Rauber's layer (Fig. 3F). Serial histological sections verified that OCT4 was mainly expressed in epiblast cells and a few Rauber's layer cells, except for the hypoblast (Fig. 3f1, f2). In the embryos that Rauber's layer had degenerated (Fig. 3G), OCT4 was only found in the nuclei of epiblast cells (Fig. 3g1, g2).
VIMENTIN expression indicates mesoderm differentiation and migration
VIMENTIN is a marker of mesenchymal cells and the mesoderm in the mouse. In the current study, the embryo proper was detached from the extra-embryonic tissue and immunohistochemical staining using anti-VIMENTIN and anti-OCT4 antibodies was employed to identify mesoderm differentiation and migration during porcine embryonic development. Whole mount staining with the anti-VIMENTIN antibody showed positive cells in the embryo proper, and some positive cells in the trophoblast, which was located posterior and lateral to the embryo proper (Fig. 4A). Serial sections by transverse sectioning revealed that VIMENTIN was highly expressed in the cell membrane of epiblast cells (Fig. 4a1), and moderately expressed in the membrane of cells located under the epiblast and bilateral along the embryonic disc borders (Fig. 4a2, a3). Moreover, the primitive groove (Fig. 4a3, a4), and thickened area of epiblast were observed at one end of the unfolding process (Fig. 4a5). Again, VIMENTIN was noticeably expressed in cells of the thickened area (Fig. 4a5).
Expression of VIMENTIN indicates mesoderm differentiation. The embryo proper was detached from the extra-embryonic tissue, and immunohistochemically stained for VIMENTIN and OCT4. (A) The detached embryo proper that was whole mount stained for VIMENTIN. (a1–a5) Representative serial sections of the whole mount-stained embryo proper. VIMENTIN was present in the membrane of extra-embryonic mesoderm cells and others cells slitting off at the caudal and bilateral along the embryonic disc borders. In addition, VIMENTIN was noticeably present in mesoderm and epiblast cell membranes. (B) The detached embryo proper that was whole mount stained for OCT4. (b1, b2) Representative serial sections of the whole mount-stained embryo proper. OCT4 was mainly located in the nuclei of primitive streak and epiblast cells, but weakly expressed in migrating mesoderm cells. No Oct4 expression was found in the endoderm or trophoblast. The position and orientation of the sections are indicated with white bars. Insets showed larger magnification of boxed regions. P to A: posterior–anterior axis. Exm: extra-embryonic mesoderm; Me: mesoderm; PG: primitive groove; TE: trophoblast; VE: visceral endoderm.
At the stage of mesoderm differentiation, OCT4 was also detected in embryo proper detached from the conceptus at around D11 (Fig. 4B). Representative serial sections along the anterior–posterior axis revealed that the primitive streak and epiblast nuclei were intensely stained for OCT4, migrating mesoderm cells were weakly stained, and endoderm and trophoblast cells were negatively stained (Fig. 4b1, 2).
Discussion
The current study presents detailed morphological findings on porcine embryos at D7–D13 post-insemination. The results showed apparently asyn-chronous development of the embryos that exhibited discrepancies in their shape and size within and between litters, although they were collected at the same number of days following insemination. There was considerable variability in the morphology of embryos within the developmental period, including spherical, ovoid, tubular and filamentous morphologies. This phenomenon has been commented on in several previous studies (Patten, Reference Patten1948; Anderson, Reference Anderson1978). To identify further characterization of the changes of morphology, the processes of embryo proper differentiation and germ layer-specific molecular marker expression were investigated.
The first lineage segregation occurs during the morula to blastocyst stage, which is around day 5 in porcine embryos, and results in two populations of cells, namely the ICM and TE (Perry & Rowlands, Reference Perry and Rowlands1962). At D7, the current data showed that all embryos were devoid of ZP, which was consistent with previous report (Barends et al., Reference Barends, Stroband, Taverne, Kronnie, Leën and Blommers1989). After hatching from the ZP, the first evidently fundamental event in embryonic development is formation of the embryonic disc. The current results demonstrated that porcine embryos underwent sequential morphological changes during this process. First, the polar trophoblast became obviously distinguished from the ICM, which is defined as Rauber's layer. The hypoblast migrated from the ICM and lined the inside of the TE, which is called PE, and the remaining cells in the ICM are known as the epiblast at this stage, which resembled to structure of D8 by the description of Barends et al. (Reference Barends, Stroband, Taverne, Kronnie, Leën and Blommers1989). Furthermore, the hypoblast underlying the epiblast showed the characteristic of closely packed simple epithelial cells and asymmetrical distribution (Hassoun et al., Reference Hassoun, Schwartz, Feistel, Blum and Viebahn2009), which were similar to the cellular architecture of the mouse anterior visceral endoderm (Joyce & Srinivas, Reference Joyce, Srinivas and Kubiak2012). These characteristics were considered as the primary indications that anterior differentiation of the hypoblast had begun. Next, a small cavity appeared in the epiblast, which was surrounded by developing epiblast cells, except for an opening covered by Rauber's layer. As development proceeded, Rauber's layer was gradually lost and the cavity opened, resulting in direct contact of the epiblast with the external environment, all of which formed the so-called embryonic disc. At this point of development, the epiblast became composed of tall, columnar epithelial-like cells. Previous data suggested that the two pre-streak stages of Vejlsted might be useful to subdivide Hassoun stage 2 into an early (Rauber's layer removed) and late (pseudostratified posterior epiblast) stage 2. Compared with the previous data, the embryonic disc formation that was observed in the current study was consistent with the developmental stages described by Vejlsted et al. (Reference Vejlsted, Offenberg, Thorup and Maddox-Hyttel2006a,b) and Hassoun et al. (Reference Hassoun, Schwartz, Feistel, Blum and Viebahn2009), which was named as pre-streak stage 1 or the early stage 2 (Rauber's layer removed). However, the most interesting finding in the current study was the cavity inside of the epiblast prior to formation of the embryonic disc. There are some reports (Barends et al., Reference Barends, Stroband, Taverne, Kronnie, Leën and Blommers1989; van de Pavert et al., Reference van de Pavert, Schipper, de Wit, Soede, van den Hurk, Taverne, Boerjan and Stroband2001; Hall et al., Reference Hall, Jacobsen, Rasmussen and Hyttel2010) that indicated that there is a space in the epiblast before degeneration of Rauber's layer. In the study, three out of nine embryos showed that a cavity existed in epiblast. The images of typical specimens also showed indisputably the existence of a distinct cavity inside the epiblast, which was overlaid with Rauber's layer and which unfolded step by step (Hall et al., Reference Hall, Jacobsen, Rasmussen and Hyttel2010). The current results also demonstrated that the cavity appeared in the epiblast while the hypoblast appeared taller in the presumptive anterior end of the embryonic disc during anterior differentiation. van de Pavert et al. (Reference van de Pavert, Schipper, de Wit, Soede, van den Hurk, Taverne, Boerjan and Stroband2001) discovered that a space started to appear in the centre of epiblast and the anterior–posterior differentiation began on approximately day 9 after ovulation in porcine embryos. The subsequent dramatic unfolding of the epiblast occurred on day 10 after ovulation. Therefore, the cavity may appear between stage 1 (anterior differentiation in the hypoblast) and early stage 2 (pre-streak stage 1, Rauber's layer removed), which may be a normal embryonic morphological change during early development. Compared with mouse embryonic development, the porcine embryo develops in a completely different manner, but the presence of a cavity inside of the porcine epiblast and anterior differentiation in the hypoblast indicate that some conserved features are shared by the two species. However, the developmental consequence of the cavity in porcine embryos is an opened flat embryonic disc instead of differentiation of a proamniotic cavity that occurs in mouse embryos.
After formation of the embryonic disc, the next step of development is gastrulation, characterized by the presence of the primitive streak. In the current study, it was found that the initial event was a crescent-shaped thickening that appeared at the caudal portion of the embryonic disc. Typical specimens showed that mesoderm cells appeared and migrated along the caudal portion of the primitive streak. These features were in accordance with stage 3 defined by Hassoun et al. (Reference Hassoun, Schwartz, Feistel, Blum and Viebahn2009) and pre-streak 2 devised by Vejlsted et al. (Reference Vejlsted, Offenberg, Thorup and Maddox-Hyttel2006a,b). In agreement with Hassoun et al. (Reference Hassoun, Schwartz, Feistel, Blum and Viebahn2009), the primitive node that Vejlsted et al. (Reference Vejlsted, Du, Vajta and Maddox-Hyttel2006a) described at the primitive stage was not found in the current study. It can be concluded that this discrepancy may have been caused by the time of embryo collection, and it also indicates that the time window of the primitive node in porcine embryos may be quite short.
To further confirm the observed characteristics of morphological differentiation, embryos were immunohistochemically stained for germ layer-specific molecular markers that have been well proven in mouse embryos. GATA6 (Koutsourakis et al., Reference Koutsourakis, Langeveld, Patient, Beddington and Grosveld1999; Kuijk et al., Reference Kuijk, Du Puy, Van Tol, Oei, Haagsman, Colenbrander and Roelen2008) and VIMENTIN (Evans et al., Reference Evans, Notarianni, Laurie and Moor1990; Yamada et al., Reference Yamada, Kawamata, Walker and McGeer1992) were used as markers to identify the PE and mesoderm, respectively, and OCT4 and SOX2 were used to mark pluripotent cells in the epiblast. It was found that GATA6 staining was negative in newly hatched embryos, but positive in cells lining the inside of the TE in the extra-embryonic area during embryonic disc establishment. Considering the reference information from morphological studies (Fig. 2a1,2), these cells were clearly identified as parietal endoderm. Therefore, GATA6 can be confidently used as a PE marker in porcine embryos for future studies. In addition, combined with unpublished data from the current work, the expression pattern of GATA6 in porcine embryos from hatching to the gastrulation stage closely resembled that in mouse embryos (Koutsourakis et al., Reference Koutsourakis, Langeveld, Patient, Beddington and Grosveld1999; Cai et al., Reference Cai, Capo-Chichi, Rula, Yang and Xu2008), which indicates GATA6 in embryos plays conserved roles between the mouse and pig.
For OCT4, the current data showed that it existed in ICM and TE of newly hatched embryo in pig, which was similar to that in early blastocyst (Kirchhof et al., Reference Kirchhof, Carnwath, Lemme, Anastassiadis, Schöler and Niemann2000; Kuijk et al., Reference Kuijk, Du Puy, Van Tol, Oei, Haagsman, Colenbrander and Roelen2008) and hatched embryos of description by du Puy et al. (Reference du Puy, Lopes, Haagsman and Roelen2011). During embryonic disc formation, the current results demonstrated by whole mount immunohistochemistry and sections that OCT4 was gradually expressed in the nuclei of epiblast cells, which is in accordance with some reports (du Puy et al., Reference du Puy, Lopes, Haagsman and Roelen2011; Wolf et al., Reference Wolf, Serup and Hyttel2011b). Also, it is decreasing when the epiblast differentiates into the mesoderm (Vejlsted et al., Reference Vejlsted, Offenberg, Thorup and Maddox-Hyttel2006b; Wolf et al., Reference Wolf, Serup and Hyttel2011b). The results of the current study supported these observations. Unlike OCT4, transcription factor SOX2 has received less attention in previous reports of porcine embryos. The current study found that SOX2 was expressed in the ICM of hatched porcine blastocysts and in the epiblast during formation of the embryonic disc, and unpublished data also show specific SOX2 expression in the ICM of early blastocysts up to D6. The expression pattern of SOX2 revealed by the current study did not support the description of Hall et al. (Reference Hall, Jacobsen, Rasmussen and Hyttel2010) and du Puy et al. (Reference du Puy, Lopes, Haagsman and Roelen2011), who suggested that SOX2 but not mRNA is only localized in the ICM of D8.5 embryos and epiblast. Taken together, OCT4 and SOX2 can be used as a pluripotency marker for epiblast cells in the porcine embryo.
VIMENTIN-staining results indicated that VIMENTIN was expressed in cells located in the posterior of the embryonic disc, the primitive streak, and migrating cells underlying the embryonic disc as well as epiblast cells located in the anterior primitive streak. Combined with the results of OCT4 expression, which showed decreasing expression in the posterior embryonic disc, primitive streak, and migrating cells, it has been further confirmed that VIMENTIN is an appropriate marker for labelling mesoderm in porcine embryos. Moreover, together with serial sections of embryos stained by whole mount immunohistochemistry, a mesoderm differentiation pattern revealed that ingressive movement occurred at the posterior of the embryonic disc and migrated bilaterally along the embryonic disc borders. This result was consistent with the pattern of mesoderm differentiation revealed by tracing Brachyury (T) as a mesoderm marker in a study by Wolf et al. (Reference Wolf, Serup and Hyttel2011c). In general, a mesoderm differentiation pattern underwent ingressive movement at the posterior of the embryonic disc and migrated bilaterally along the embryonic disc borders.
In conclusion, hypoblast and mesoderm differentiation was demonstrated in in vivo porcine preimplantation embryos (D7–D13). During embryonic disc formation, the hypoblast showed characteristics similar to those of the mouse anterior visceral endoderm. In addition, a cavity appeared in the epiblast, which was similar to a mouse proamniotic cavity. The cavity opened and directly contacted the external environment, and formed the embryonic disc ultimately. Furthermore, GATA6 can clearly mark parietal endoderm cells during embryonic disc establishment. Unlike OCT4, SOX2 was expressed in the ICM of hatched blastocysts and the epiblast during formation of the embryonic disc. Therefore, both SOX2 and OCT4 can be used as markers of pluripotent cells in the porcine embryonic disc. In addition, at the start of gastrulation, VIMENTIN, OCT4 analysis and serial sections of embryos stained revealed that the mesoderm differentiation pattern underwent ingressive movement at the posterior of the embryonic disc and migrated bilaterally along the embryonic disc borders.
Acknowledgments
This work was supported by the National Program on Key Basic Research Project (973 Program; Grant Number: 2011CB944202; is displaced with 2012CBA01303).
