The use of ART (assisted reproductive techniques) for the treatment of human infertility/subfertility is rapidly increasing; according to estimates, in developed countries 1-2% of children are born via ART [1, 2]. However, since the timing of ART coincides with that of key epigenetic events, epigenetic regulation may be susceptible to disruption. In particular, ovarian stimulation and in vitro culture have a high risk of DNA methylation disruption .
DNA methylation, the most characterized epigenetic modification, is involved in transcriptional repression, global X chromosome inactivation and genomic imprinting . DNA methylation is catalyzed by a family of DNA methyltransferases (DNMTs). In mammals, five DNMTs (DNMT1, DNMT2, DNMT3A, DNMT3A and DNMT3L) have been defined . They have two types of methyltransferase activity: de novo and maintenance methylation. DNMT3A and DNMT3B, the de novo DNMTs, play important roles in the establishment of DNA methylation patterns, but are dispensable for the maintenance of the methylation marks at most imprinted loci during preimplantation development . DNMT1 (DNMT1s and DNMT1o being the somatic and oocyte-specific isoforms of DNMT1, respectively), a major maintenance DNMT, is essential for methylation maintenance. In mice, Dnmt1 alone is sufficient for the maintenance of methylation marks of imprinted genes during preimplantation development .
Methylation patterns required for genomic imprinting change dynamically during mammalian development. In general, imprinting is thought to be a multi-step process involving erasure, establishment and maintenance of a methylation mark. In the mouse, the paternal genome undergoes active global demethylation within 6–8 h of fertilization while passive demethylation of the maternal genome occurs during the cleavage stage [7, 8]; however, imprinted genes and some repeated sequences maintain methylation during this wave of global demethylation . Imprints are erased in primordial germ cells and must then be re-established during gametogenesis in a sex-specific manner [10, 11]. During oogenesis, maternal imprints are acquired at a specific time for each gene during postnatal oocyte growth. This process was also found to correlate with an increase in oocyte diameter [12, 13]. During spermatogenesis, paternal imprints are established in the postnatal stage [14, 15]. If methylated imprints are incorrectly established during gametogenesis, these defects may be found in the resulting pregnancy.
Studies have been carried out to determine the effect of ovarian stimulation on DNA methylation during oogenesis, preimplantation and postimplantation development . Interestingly, the results reported are conflicting. For oocytes, Sato et al.  demonstrated a gain in H19 methylation in mouse oocytes and in human oocytes derived from superovulated females, but a loss of PEG1 methylation in the latter. In contrast, Anckaert et al.  found that superovulation did not affect methylated imprint acquisition at H19, Snrpn or Igf2r in mouse oocytes. Moreover, a recent study showed that H19, Snrpn, Peg3 and Kcnq1ot1 have normal methylated imprint patterns in mouse oocytes when low- or high-dose hormone was administered . During preimplantation development, H19 methylation was unaltered, whereas in mouse blastocysts H19 expression was disturbed ; however, loss of methylation at the maternally imprinted genes Snrpn, Peg3 and Kcnq1ot1 and gain of methylation at the paternally imprinted gene H19 were observed in blastocysts by Market-Velker et al. . Similarly, global methylation in 2-cell stage embryos derived from superovulated mice was two times higher than that in non-superovulated counterparts . During postimplantation development, superovulation resulted in biallelic expression of Snrpn and H19 at 9.5dpc (day post coitus) placentas while Kcnq1ot1expression was not affected . Interestingly, DNA methylation at Snrpn and H19 was unaltered . Taken together, these data show that the effect of superovulation on DNA methylation varies and is incompletely understood.
Ovarian stimulation was found to disturb DNA methylation at not only maternally but also paternally imprinted genes in mouse blastocysts . Such dual effects on methylated imprints may originate from abnormal imprint acquisition or from disruption of maternal-effect gene production, which is required for later imprint maintenance. Abnormal methylated imprint acquisition cannot explain methylation defects in blastocysts since methylated imprints are correctly established in oocytes . Therefore, defects in imprint maintenance in blastocyst are suspected to originate from the disruption of gene products stored in oocytes.
Long interspersed elements-1 (LINE-1) is highly repeated human retrotransposon elements found in large numbers in eukaryotic genomes . LINE-1s constitute about 17% of human genome as 600,000 copies and about 3000–400 copies of LINE-1s remains as full length form and some of them may retain its activity . The intracisternal A-particle (IAP) element is a long terminal repeat (LTR)-type mouse retrotransposon, which is consisted with gag, pro, and pol genes . Recently it was previously reported that the most extreme methylation changes during the sperm to zygote transition are enriched for LINEs . Especially LINE L1 is the one with most significant decrease in methylation level, while class II intercisternal A-particles (IAP) does not show any methylation within zygotes. LINE and LTR activity in the early embryo is associated with some of the earliest transcriptional events during zygotic genome activation .
The objective of the present study was to further investigate the effect of superovulation on methylation and to determine whether the expression of Dnmts, which is required for methylation acquisition and maintenance, is disturbed upon ovarian stimulation. Two hormone dosages, low (6 IU) or high (10 IU), were administered. Dnmt mRNA levels were assessed in metaphase II (MII) oocytes and zygotes. In addition, to further understand the effect of superovulation on methylation, global DNA methylation and H3K9 methylation were investigated in zygotes, and DNA methylation was assessed at repeated sequence (IAP LTR and Line1 5′)in blastocyst stage embryos.