br Materials and methods br Results br Discussion

Materials and methods
Results
Discussion The roles of EZH2 during oocyte maturation, early embryogenesis, and somatic cell nucleus transfer (SCNT) have been studied extensively in mouse [12,30]. How it functions during porcine preimplantation development, especially the maternal effect, however, remains poorly understood. Using siRNA interference, we demonstrated that EZH2 knockdown disrupted the equilibrium between differentiation and pluripotency of cells in the pig parthenogenetic embryo, suggesting that EZH2 is integral to its developmental program. During pig oocyte maturation and preimplantation development, EZH1 and EZH2 both underwent dynamic changes. However, only that of EZH2 was accompanied by changes in global H3K27me3 levels (Fig. 1), suggesting a correlation between EZH2 and H3K27me3. Despite this correlation, different results have been obtained regarding the dynamics of both EZH2 and H3K27me3. In our study, H3K27me3 was persistently detectable during the whole process of oocyte maturation and at the PN stage but barely detectable in cleavage stage embryos (from 2-cell to compacted morula stage). However, at the SJB3-019A stage, our results showed that H3K27me3 became detectable again, which is consistent with that reported by Gao et al. [31]. But there is also a study reporting the missing of H3K27me3 staining at the blastocyst stage in porcine [32]. A few reasons may contribute to the apparent differences in the staining patterns of H3K27me3 in porcine blastocysts. Firstly, it may be due to the difference in the compared developmental stages and the different time points when the embryos were collected. Both Gao et al. and Park et al. have performed H3k27me3 staining with in vitro fertilization (IVF) embryos. The later found that the staining was from faint to absent in blastocysts, while the former performed staining on hatched blastocysts, which develop further than blastocysts, and showed that H3K27me3 was higher in the trophectoderm than in the epiblast. Because of the lower ratio of ICM cells/total cells in parthenogenetic blastocysts than that in IVF embryos [33], this might make it relatively easier for us to detect H3K27me3 in parthenogenetic embryos. In particular, there are also inconsistent reports regarding the staining patterns of the H3K27me3 mark in the blastocysts of murine. Wu and Zhang et al. found that the intensity of H3K27me3 staining in blastocysts was weaker than that in morula [18,34]. Erhardt et al. observed a weak overall staining in trophectoderm cells except in some special areas, whereas the ICM showed an intense staining [35]. Interestingly, Yang et al. showed that H3K27me3 modification distributed with high levels in both the ICM and the trophoblast cells in early blastocysts, and the distribution was more extensive than that in morula. Also, they found that the H3K27me3 mainly aggregated in the ICM of hatched blastocysts [36]. Secondly, there may be differences in the sensitivity of the experimental methods used. The global level of H3K27me3 in pig blastocysts might just below the detectable level by some methods. We speculated that H3K27me3 would increase by at least the blastocyst stage since it plays a crucial role in X-chromosome inactivation. Furthermore, our results about the H3K27me3 staining in porcine blastocysts are similar to the observations of Huang et al. [37]. Besides, the abundance of H3K27me3 in parthenogenetic blastocysts has also been detected by Western blot in our study (Fig. 4E). Thirdly, these discrepancies might originate from different antibody accessibility. Trimethylated H3K27 was surrounded by higher-order chromatin structures, which have the potential to disturb the process of antibody accessing the antigen, especially when different antibodies were used in these reports. Similarly, different results have also been reported for the expression pattern of EZH2 during porcine preimplantation development. For example, Gao et al. showed a steady increase of EZH2 from the 4-cell to the hatched blastocyst stage, but Park et al. showed a reduction in EZH2 expression during the same period of embryonic development. This difference is probably due to differences in the origin of embryos, the culture conditions, the sample processing and the reference genes. Interestingly, the activity of Ezh2 in mouse is also controversial. Some reported that it was already apparent in late stage zygotes [38,39], but Wu et al. showed that Ezh2 expression peaked during the zygote stage but then gradually decreased from the 2-cell stage and reached the lowest level at the blastocyst stage [18]. These findings above indicate that the establishment and maintenance of global H3K27me3 requires EZH2, which is regulated by a potential mechanism in different mammals before implantation.