Introduction
Normal embryo development involves important rearrangements in the chromatin structure. This change includes a global reprogramming of epigenetic marks starting following fertilization, which is required to create a totipotent state zygote (Fraser & Lin, Reference Fraser and Lin2016). Epigenetic marks are essential elements of the cell machinery that functionally interprets DNA sequences. These marks contribute to the establishment of cell identity by regulating gene expression patterns that are specific for each cell lineage (Chen & Pei, Reference Chen and Pei2016). In addition to regulate gene expression, epigenetic marks are also involved in the regulation of other critical events during embryogenesis in mammals, including genomic imprinting, inactivation of the X chromosome and reprogramming of the two parental haploid genomes into one diploid genome (Morris, Reference Morris2009; Hales et al., Reference Hales, Grenier, Lalancette and Robaire2011).
Important epigenetic marks involved in the regulation of early embryo development include DNA methylation, non-coding RNAs and post-translational modifications of specific residues in nucleosomal histones, such as acetylation and methylation (Novina & Sharp, Reference Novina and Sharp2004; Cook & Blelloch, Reference Cook and Blelloch2013; Beaujean, Reference Beaujean2014; Dallaire & Simard, Reference Dallaire and Simard2016; Ambrosi et al., Reference Ambrosi, Manzo and Baubec2017). Histone acetylation usually occurs on the lysine residues of core histones and neutralizes the basic charge of these residues, therefore decreasing their affinity for DNA (Hasan & Hottiger, Reference Hasan and Hottiger2002). Histone acetylation is almost invariably associated with activation of gene transcription, and is involved in the regulation of cell totipotency and proliferation events during cell reprogramming and embryo development (Wang et al., Reference Wang, Kou, Zhang and Gao2007; Rodriguez-Sanz et al., Reference Rodriguez-Sanz, Moreno-Romero, Solis, Kohler, Risueno and Testillano2014). For instance, acetylation of histone H3 at lysine 27 (H3K27ac) plays an important role in the maintenance of pluripotency and regulation of key developmental genes in stem cells (Creyghton et al., Reference Creyghton, Cheng, Welstead, Kooistra, Carey, Steine, Hanna, Lodato, Frampton and Sharp2010; Pasini et al., Reference Pasini, Malatesta, Jung, Walfridsson, Willer, Olsson, Skotte, Wutz, Porse and Jensen2010).
Conversely, methylation of H3K27 (H3K27me) is usually associated with transcriptional repression, promoting stable and heritable gene silencing (Schwartz & Pirrotta, Reference Schwartz and Pirrotta2007). Trimethylation of H3K27 (H3K27me3) regulates lineage specification by temporarily repressing genes involved in development and cell differentiation of pluripotent cells (Surface et al., Reference Surface, Thornton and Boyer2010; Shpargel et al., Reference Shpargel, Starmer, Yee, Pohlers and Magnuson2014). Indeed, genes associated to organogenesis, morphogenesis and embryonic development are temporarily suppressed by H3K27me3 until the time when their transcription is required (Boyer et al., Reference Boyer, Plath, Zeitlinger, Brambrink, Medeiros, Lee, Levine, Wernig, Tajonar and Ray2006).
It is well established that mammalian embryos undergo several epigenetic modifications during early development (Boland et al., Reference Boland, Nazor and Loring2014). Nonetheless, the regulatory mechanism of lineage commitment is still not fully known. Understanding how different species organize their chromatin following fertilization is fundamental for comprehending the epigenetic regulation of early embryonic development and may help to determine improved conditions for in vitro embryo production. Therefore, the objective of this study was to evaluate the profile of acetylation and methylation of histone H3 at lysine 27 during oocyte maturation and early development in vitro of porcine embryos.
Materials and Methods
Unless otherwise indicated, chemicals were purchased from Sigma Chemical Company (Sigma-Aldrich, Oakville, ON, Canada).
Oocyte collection and in vitro maturation
Ovaries from prepubertal gilts were collected from a local abattoir and transported to the laboratory in 0.9% NaCl at 30 to 35°C. Cumulus–oocyte complexes (COCs) were aspirated from 3-mm to 6-mm diameter follicles using an 18-gauge needle. Only COCs surrounded by a minimum of three cumulus-cell layers, with an evenly granulated cytoplasm were selected for in vitro maturation (IVM). Groups of 20 to 25 COCs were cultured in 100 µl of maturation medium under mineral oil, in a humidified atmosphere of 5% CO2 and 95% air at 38.5°C. Maturation medium consisted of TCM-199 (Life Technologies, Burlington, ON, Canada), supplemented with 0.1 mg/ml cysteine, 0.91 mM sodium pyruvate, 3.05 mM d-glucose, 0.5 µg/ml follicle-stimulating hormone (FSH; SIOUX Biochemical Inc.), 0.5 µg/ml luteinizing hormone (LH, SIOUX Biochemical Inc.), 10 ng/ml epidermal growth factor (EGF; Life Technologies), 20 µg/ml gentamicin (Life Technologies), 1 mM dibutyryl cyclic adenosine monophosphate (dbcAMP), and 20% (v/v) porcine follicular fluid. After 22 h, the oocytes were transferred to the same maturation medium, but without FSH, LH, and dbcAMP for an additional 22 to 24 h, under the same conditions.
In vitro fertilization (IVF) and in vitro culture (IVC)
Matured oocytes were freed from cumulus cells by vortexing in TCM 199 HEPES-buffered medium (Life Technologies) supplemented with 0.1% hyaluronidase for 7 min. The denuded oocytes were washed in the same medium before being transferred to fertilization medium.
The oocytes were washed twice in fertilization medium, which consisted of Tris-buffered medium (TBM) supplemented with 2.5 mg/ml fatty acid-free bovine serum albumin (BSA), and placed in 90 µl drops of the same medium. A pool of fresh semen collected from 3–4 fertile boars was left at 18°C for 24 h and then centrifuged at 4000 rpm for 3 min. Supernatant was discarded and the sperm pellet was re-suspended in 1 ml of fertilization medium and homogenized. The semen was centrifuged again (4000 rpm for 2 min), the supernatant was discarded and the sperm pellet was re-suspended in 500 µl of fertilization medium. Sperm concentration was adjusted to obtain a final concentration of 2,000 live sperm/oocyte. Sperm and oocytes were co-incubated for 4 h under mineral oil at 38.5°C, in 5% CO2 in air and 100% humidity.
After IVF, the oocytes were washed three times in porcine zygote medium (PZM-3) supplemented with 3 mg/ml BSA and then cultured in the same medium under mineral oil. Culture conditions were humidified atmosphere of 5% CO2 and 95% air at 38.5°C. Cleavage and blastocyst rates were determined at 48 h and 7 days after IVF, respectively.
Immunocytochemistry
Oocytes/embryos were fixed at the following developmental stages: germinal vesicle (GV) and metaphase II (MII) oocytes, pronuclear (PN; 18 hours post fertilization – hpf), 2-cell stage (2C; 36 hpf), 4-cell stage (4C; 48 hpf), 8-cell stage (8C; 72 hpf), D6 blastocysts (D6; 144 hpf) and D8 blastocysts (D8; 192 hpf). Samples were rinsed in PBS, fixed in 4% paraformaldehyde for 15 to 20 min and stored at 4°C in PBS with 0.2% TritonX-100 and 0.3% BSA. Cell permeabilization was performed with 0.5% Triton X-100 in PBS with 0.3% BSA for 1 h at 37°C. Oocytes/embryos were then washed twice (10 min each) in blocking solution (3% BSA and 0.2% Tween-20 in PBS) and exposed overnight at 4°C to primary antibodies diluted in blocking solution (1:1000). Polyclonal rabbit anti-H3K27 acetyl (Abcam; ab4729), anti-monomethyl H3K27 (Upstate; 07–448), anti-dimethyl H3K27 (Upstate; 07–452), and anti-trimethyl H3K27 (Upstate; 07–449) were used as primary antibodies. Samples were then washed three times for 20 min each in blocking solution and incubated for 2 h at room temperature in the presence of 1:1000 diluted Alexa Flour 488 goat anti-rabbit (Molecular Probes, Eugene, OR, USA) secondary antibodies. Finally, the samples were washed three times (20 min each) in blocking solution. For the second wash, the solution was supplemented with 10 µg/ml DAPI (4´,6-diamidino-2-phenylindole) for DNA staining. Oocytes/embryos were mounted on microscope slides using a drop of Mowiol and examined by epifluorescence using a Nikon eclipse 80i microscope (Nikon, Tokyo, Japan) with ×200 magnification. Images were individually recorded using a Retiga 2000R monochrome digital camera (Qimaging, BC, Canada). The exposure gains and rates were consistent between samples. Fluorescence intensities were quantified by using Image J analysis (Schneider et al., Reference Schneider, Rasband and Eliceiri2012). Control samples from each developmental stage were processed as described above, but the primary antibody was omitted. Images from 183 oocytes/embryos were analyzed.
Results
H3K27ac was detected in the nuclei of GV and in the DNA of MII oocytes. Strong fluorescent signals were observed in both PN and in 2C and 4C stages. The fluorescent signal decreased at the 8C stage but increased again in D6 and D8 blastocysts (Figs 1 and 2).
Figure 1 Distribution of acetylated, mono-, di- and tri-methylated H3K27 in porcine oocytes and embryos. Panels show representative images of immature (GV stage) and mature (MII stage) oocytes, zygotes (PN stage), 2-cell (2C), 4-cell (4C) and 8-cell (8C) stage embryos, and day 6 (D6) and day 8 (D8) blastocysts.
Figure 2 Fluorescence intensity for H3K27ac (A), H3K27me1 (B), H3K27me2 (C) and H3K27me3 (D) in porcine oocytes and developing embryos. GV = germinal vesicle oocytes, MII = metaphase II oocytes, PN = pronuclear stage zygotes, 2C = two-cell stage embryos, 4C = four-cell stage embryos, 8C = eight-cell stage embryos, D6 = day 6 blastocysts, D8 = day 8 blastocysts.
The fluorescent signal for H3K27me1 was intense in GV and MII oocytes, but decreased markedly in cleaved-stage embryos. The fluorescent signal was barely detectable in zygotes, 2C, 4C and 8C embryos, but intensity increased in D6 and D8 blastocysts (Figs 1 and 2).
H3K27me2 revealed a strong fluorescent signal in GV and MII stage oocytes, but very weak or absent signal in PN of zygotes. Cleaved-stage embryos showed weak or no fluorescent signal for H3K27me2, but all nuclei were H3K27me2 positive in D6 and D8 blastocysts (Figs 1 and 2).
Strong fluorescence intensity for H3K27me3 was detected in GV and MII oocytes. In zygotes, only one of the pronuclei was H3K27me3 positive (Fig. 3). H3K27me3 signal was detected in 2C and 4C stages, but 8C embryos and D6 blastocysts showed very weak or no fluorescent signal (Figs 1 and 2). The fluorescent pattern in D8 blastocyst varied from absence of fluorescent signal (7 embryos), mixture of fluorescent positive and negative cells (19 embryos), to only inner cell mass (ICM) positive staining (three embryos; Fig. 4).
Figure 3 H3K27me3 immunofluorescence image of a polyspermic zygote showing one stained and two non-stained PNs. (1) DNA stained with DAPI. (2) H3K27me3 fluorescent signal.
Figure 4 H3K27me3 immunofluorescence image day 8 blastocysts. (1) Embryo with absence or weak H3K27me3 fluorescent signal. (2) Embryo with strong H3K27me3 fluorescent signal in all the cells. (3) Embryo with strong H3K27me3 fluorescent signal only in the ICM cells.
Discussion
Epigenetic changes are known to regulate the programming of the genetic information carried over by the oocyte and the sperm to the new developing embryo and control totipotency and cell lineage commitment (Lu & Zhang, Reference Lu and Zhang2015; Ancelin et al., Reference Ancelin, Syx, Borensztein, Ranisavljevic, Vassilev, Briseno-Roa, Liu, Metzger, Servant and Barillot2016). To gain additional insights into the epigenetics changes that occur at early embryo stages we have evaluated the global profile of H3K27 acetylation, mono-, di-, and trimethylation during in vitro development of porcine embryos. H3K27 methylation is catalyzed by the transcriptional repressors Polycomb group (PcG) proteins, which form two chromatin modifying complexes, Polycomb Repressive Complex 1 and Complex 2 (PRC1 and PRC2) (O'Meara & Simon, Reference O'Meara and Simon2012; Williams et al., Reference Williams, Christensen, Rappsilber, Nielsen, Johansen and Helin2014). Removal of this epigenetic mark can make undifferentiated normal (Patel et al., Reference Patel, Tursun, Rahe and Hobert2012) or cancerous cells (Gannon et al., Reference Gannon, De Long, Endo-Munoz, Hazar-Rethinam and Saunders2013; Ciarapica et al., Reference Ciarapica, Carcarino, Adesso, De Salvo, Bracaglia, Leoncini, Dall'agnese, Verginelli, Milano and Boldrini2014) susceptible to differentiation. H3K27me3 is a Polycomb repressed state, which is functionally opposed by actively transcribed chromatin (O'Meara & Simon, Reference O'Meara and Simon2012). Conversely, H3K27ac is generally enriched in promoters of active genes (Tie et al., Reference Tie, Banerjee, Stratton, Prasad-Sinha, Stepanik, Zlobin, Diaz, Scacheri and Harte2009; O'Meara & Simon, Reference O'Meara and Simon2012) and can antagonize PcG activity by competing with the placement of the H3K27me3 mark (Schwartz et al., Reference Schwartz, Kahn, Stenberg, Ohno, Bourgon and Pirrotta2010). Consistent with this, it has been shown that H3K27ac and H3K27me3 have dynamic and complementary temporal profiles during embryogenesis (Tie et al., Reference Tie, Banerjee, Stratton, Prasad-Sinha, Stepanik, Zlobin, Diaz, Scacheri and Harte2009).
In this study, H3K27me3 was detected in immature and matured oocytes, which is in accordance with others studies in pigs (Park et al., Reference Park, Magnani and Cabot2009; Cao et al., Reference Cao, Li, Chen, Wang, Zhang, Zhou, Wu, Ling, Fang and Li2015; Xie et al., Reference Xie, Zhang, Wei, Li, Weng, Kong and Liu2016) and cattle (Ross et al., Reference Ross, Ragina, Rodriguez, Iager, Siripattarapravat, Lopez-Corrales and Cibelli2008). In zygotes, only one of the two pronuclei was positive for this mark. A similar pattern of H3K27me3 staining in zygotes was reported in other studies in pigs (Young et al., Reference Young, Yeo, Jung, Lee and Kang2007; Park et al., Reference Park, Magnani and Cabot2009), cattle (Breton et al., Reference Breton, LE Bourhis, Audouard, Vignon and Lelièvre2010) and mice (Erhardt et al., Reference Erhardt, Su, Schneider, Barton, Bannister, Perez-Burgos, Jenuwein, Kouzarides, Tarakhovsky and Surani2003; Santos et al., Reference Santos, Peters, Otte, Reik and Dean2005). We believe that the H3K27me3 positive was the female PN given that polyspermic zygotes having more than two PNs also contained only one positive PN. However, Ross et al. (Reference Ross, Ragina, Rodriguez, Iager, Siripattarapravat, Lopez-Corrales and Cibelli2008) observed an asymmetric staining pattern in bovine embryos at PN stage and proposed that the female PN was the one positive for H3K27me3. They have also observed that both pronuclei in parthenogenetically activated embryos were H3K27me3 positive, which further confirms that the positive PN in fertilized zygotes is derived from the oocyte.
We have observed that the fluorescent intensity for H3K27me3 decreases at the 8-cell stage. This finding is in line with other reports in porcine (Park et al., Reference Park, Magnani and Cabot2009) and bovine embryos (Ross et al., Reference Ross, Ragina, Rodriguez, Iager, Siripattarapravat, Lopez-Corrales and Cibelli2008; Breton et al., Reference Breton, LE Bourhis, Audouard, Vignon and Lelièvre2010). Interestingly, we observed different patterns of H3K27me3 fluorescent staining in D8 blastocysts, which included embryos with no stained cells, embryos with a mix of positive and negative cells, embryos with all the cells stained, and embryos with only the ICM stained. A previous study by Park et al. (Reference Park, Magnani and Cabot2009) did not detect H3K27me3 in porcine blastocysts at D6 of development (144 hpf). Conversely, Gao et al., (Reference Gao, Hyttel and Hall2010) reported that levels of H3K27me3 increased in hatched blastocysts. Along with our observations, data from those studies suggest that H3K27me3 pattern in porcine blastocysts may differ from other species given that intense fluorescence signal for H3K27me3 was found in bovine (Ross et al., Reference Ross, Ragina, Rodriguez, Iager, Siripattarapravat, Lopez-Corrales and Cibelli2008) and murine (Erhardt et al., Reference Erhardt, Su, Schneider, Barton, Bannister, Perez-Burgos, Jenuwein, Kouzarides, Tarakhovsky and Surani2003) blastocysts.
Monomethylated H3K27 was detectable in GV and MII stage oocytes, which is in agreement with a previous study in porcine embryos (Park et al., Reference Park, Magnani and Cabot2009). However, we observed that H3K27me1 was present in only one PN, while similar fluorescent signal in both PNs was reported in the previous study (Park et al., Reference Park, Magnani and Cabot2009). As H3K27me1 becomes detectable in the male pronucleus several hours after fertilization, one possibility is that the pronuclear development stage of zygotes evaluated in our study was different from the previous study. In mice, H3K27me1 signal is mainly present in the female PN (Erhardt et al., Reference Erhardt, Su, Schneider, Barton, Bannister, Perez-Burgos, Jenuwein, Kouzarides, Tarakhovsky and Surani2003; Van Der Heijden et al., Reference Van Der Heijden, Dieker, Derijck, Muller, Berden, Braat, Van Der Vlag and De Boer2005), but the signal increased during male PN formation (Erhardt et al., Reference Erhardt, Su, Schneider, Barton, Bannister, Perez-Burgos, Jenuwein, Kouzarides, Tarakhovsky and Surani2003; Santos et al., Reference Santos, Peters, Otte, Reik and Dean2005; Van Der Heijden et al., Reference Van Der Heijden, Dieker, Derijck, Muller, Berden, Braat, Van Der Vlag and De Boer2005). The H3K27me1signal from 2C to the blastocyst stage observed in this study corroborates with findings from a previous study (Park et al., Reference Park, Magnani and Cabot2009).
We observed absence or weak fluorescent intensity for H3K27me2 in zygotes, which differs from previous studies in mice reporting intensive H3K27me2 signal in PNs of mice zygotes (Erhardt et al., Reference Erhardt, Su, Schneider, Barton, Bannister, Perez-Burgos, Jenuwein, Kouzarides, Tarakhovsky and Surani2003). Conversely, we observed a prominent fluorescent signal for H3K27me2 in D6 and D8 IVF blastocysts, while another study reported that H3K27me2 levels were nearly undetectable in porcine blastocysts produced by parthenogenetic activation (Huang et al., Reference Huang, Yuan, Li, Wang, Li, Pang, Wang and Ouyang2015). Differences between fertilized and parthenogenetic embryos have been reported regarding other histone modifications including H3K9ac and H3K27ac (Huang et al., Reference Huang, Yuan, Li, Wang, Li, Pang, Wang and Ouyang2015).
Intensive fluorescent signal for H3K27ac was observed in oocytes, in both PNs of zygotes, and all developmental stages from 2C to blastocyst. However, the intensity of the fluorescent signal decreased in 8C-stage embryos. A similar pattern of H3K27ac in porcine embryos was reported in a previous study by Zhou et al. (Reference Zhou, Cao, Wu, Liu, Tao, Chen, Song, Han, Li and Fang2014).
In summary, our study revealed that the global pattern of acetylation and mono-, di- and trimethylation of the H3K27 changes during in vitro development of porcine embryos. Based on published literature, our findings suggest that porcine embryos produced by IVF may differ from embryos of other species regarding the global changes in H3K27, including the pattern of H3K27me3 at the blastocyst stage and H3K27me2 in zygotes. Our findings may provide a base for further investigation of the mechanistic relevance of each epigenetic change occurring in the H3K27 for the regulation of early embryo development.
