SEARCH

SEARCH BY CITATION

Keywords:

  • assisted reproductive techniques;
  • DNA methylation;
  • epigenetics;
  • genomic imprinting

Abstract.

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Outcome of pregnancies by ARTs
  5. Epigenetics and genomic imprinting
  6. Animal studies
  7. Human studies
  8. Conclusion
  9. Conflict of interest statement
  10. Acknowledgements
  11. References

Iliadou AN, Janson PCJ, Cnattingius S (Karolinska Institutet, Stockholm; Sweden). Epigenetics and assisted reproductive technology (Review). J Intern Med 2011; 270: 414–420.

During gametogenesis, the female and male germ cells undergo a process whereby imprinting marks are erased from the genome. During the later stages of germ-cell development, the methylation marks of the female and male germ lines are re-established. A second phase of demethylation of the genome occurs at the time of fertilization, and during development of the early embryo. Assisted reproductive technology involves several steps that subject the gametes and early developing embryos to environmental stress, and this is the primary reason for an increased interest in the putative link between these techniques and imprinting disorders. Although animal studies support a link between assisted reproductive techniques (ARTs) and imprinting disorders, via altered methylation patterns, data in humans are inconsistent. Here we provide an overview of the field of epigenetics in relation to ARTs.


Introduction

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Outcome of pregnancies by ARTs
  5. Epigenetics and genomic imprinting
  6. Animal studies
  7. Human studies
  8. Conclusion
  9. Conflict of interest statement
  10. Acknowledgements
  11. References

People throughout the world can appreciate and relate to the findings of Professor Robert Edwards, the 2010 Nobel Prize winner in physiology or medicine for the development of in vitro fertilization (IVF) techniques, which have led to the births of over 4 million babies. Although initially controversial, with both financial and ethical opposition, assisted reproductive techniques (ARTs) are now generally well accepted. Currently, ARTs account for between 1% and 4% of births annually in many western countries, and the global use of these techniques is growing [1, 2].

Recent studies suggest a possible link between human ARTs and genomic imprinting disorders [3–6]. Concern for epigenetic effects of these assisted procedures in humans has arisen from the observation of an increased incidence of rare genomic imprinting diseases, such as Beckwith–Wiedemann syndrome (BWS) and Angelman syndrome (AS), in children born after the use of ART [7–9]. One of the symptoms of BWS is overgrowth in children and the disease has been associated with altered imprinting of genes, including the IGF2R gene [7, 10]. AS is a neurocognitive disorder associated with loss of methylation in imprinted gene clusters. It is interesting that most of the imprinted genes identified to date have a role in regulating pre- and/or postnatal growth [11, 12].

In this short review, we will start by presenting the reported risks of pregnancy by ARTs and then discuss the current understanding of the link between ARTs and epigenetic disorders in both animals and humans.

Outcome of pregnancies by ARTs

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Outcome of pregnancies by ARTs
  5. Epigenetics and genomic imprinting
  6. Animal studies
  7. Human studies
  8. Conclusion
  9. Conflict of interest statement
  10. Acknowledgements
  11. References

Scientific activity to investigate the possible consequences and adverse effects of ARTs has increased exponentially during the last decade. In recent years, several studies have reported both short- and long-term adverse outcomes of ARTs. The safety of the procedures and the health of children conceived by ARTs have been questioned [13]. However, little is known about whether these adverse effects are the result of ARTs per se or the consequence of parental subfertility.

Assisted reproductive techniques have been associated with pregnancy complications (pre-eclampsia and placental complications), adverse perinatal outcomes (low birth weight, preterm birth, perinatal mortality and congenital malformations) and epigenetic alternations. However, studies of adverse outcomes after the use of ARTs have been questioned, primarily because multiple births are over-represented in assisted pregnancies. Hence, it has been difficult to determine whether pregnancy complications and adverse outcomes are a result of ARTs or owing to multiple births and their associated problems. Of interest, in a Danish national cohort study, no major differences in physical health were found between twins conceived through ARTs (ART twins) and those conceived naturally (non-ART twins) [14]. It should be noted, however, that ART twins are almost exclusively dizygotic (DZ) whereas one-third of non-ART twins are monozygotic (MZ), and hence comparisons between ART and non-ART twins are not equivalent. Thus, to investigate differences in outcomes between ART and non-ART twin pregnancies, comparisons should be made after taking zygosity into account. A study of umbilical cords of DZ ART twins showed more pathological characteristics compared to cords from DZ non-ART twins [15]. Furthermore, when ART twins were compared to singletons conceived through ARTs (ART singletons), twins were more often admitted to neonatal intensive care units and had poorer speech development, even after stratification for birth weight [16, 17]. A systematic review and meta-analysis showed that, compared to non-ART twins, ART twins were more often delivered by caesarean section, born preterm and admitted to neonatal intensive care units [18]. It was also noted that mothers who had undergone intracytoplasmic sperm injection (ICSI), which involves direct injection of sperm into the oocyte, took more sick leave and were more frequently hospitalized during pregnancy [19]. No other results were significantly different between mothers with ART twin pregnancies and those with spontaneously conceived twin pregnancies.

To reduce the risks of pregnancy and perinatal complications, a single embryo transfer was introduced as standard practice in some countries in the early 2000s. Results from the many independent published reviews and meta-analyses of perinatal outcomes in ART singletons show even stronger differences between ART and non-ART singletons compared to ART and non-ART twins [20–24]. ART singletons have on average a 2-fold increased risk of perinatal mortality and preterm birth (<37 weeks), a 50% increased risk of low birth weight (<2500 g) and an even greater risk of very low birth weight (<1500 g). ART singletons are also at increased risk of caesarean delivery, being small for gestational age, admission to a neonatal intensive care unit and birth defects. The aetiology and biological mechanisms underlying these risks in singleton ART pregnancies remain essentially unresolved. However, recently a Norwegian study, comparing naturally and ART-conceived siblings, showed no significant differences in small for gestational age foetuses and preterm births, suggesting that reported differences may be owing to parents’ infertility rather than the ART procedures [25].

Register studies in Sweden investigating pregnancy outcomes and maternal and childhood morbidity have also found increased risks of low birth weight, small for gestational age foetuses (5.1% in ART singletons vs. 2.8% in the general population of singletons) and preterm births (9.6% in ART singletons vs. 5.3% in the general population of singletons) [26–33]. Adjusting for years of involuntary childlessness reduced the risks, which nevertheless remained significantly increased for low birth weight and preterm birth [34]. A recent study showed that these risks were greater in infants conceived after embryo transfer at the blastocyst stage than after two-stage cleavage, suggesting that the duration of the embryo in culture medium could be one possible explanation for the increased risks [35]. Hence, there is evidence to suggest that ART singleton foetuses may have an increased risk of low birth weight and/or preterm birth. There is a plethora of epidemiological evidence to support the foetal programming hypothesis, which states that restricted intrauterine growth has important lifelong health implications, therefore it is essential to investigate the underlying mechanisms [36]. In the light of evidence that adverse environmental conditions early in human gestation can be recorded as persistent changes in epigenetic information [37], it is of even greater importance to study and understand potential epigenetic changes related to ARTs.

Epigenetics and genomic imprinting

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Outcome of pregnancies by ARTs
  5. Epigenetics and genomic imprinting
  6. Animal studies
  7. Human studies
  8. Conclusion
  9. Conflict of interest statement
  10. Acknowledgements
  11. References

It is well known that the phenotype of an individual is not exclusively determined by the genotype. In 1942, Waddington [38] introduced the term epigenetics, defined as ‘heritable changes in gene expression that occur without any changes in gene sequence’.

Epigenetic modifications are heritable in the sense that the ‘epigenetic status’ of the chromatin is preserved during cell mitosis. There are many different types of epigenetic modifications that are known to affect expression, including changes in nucleosome positioning and conformation, and histone modifications such as acetylation, phosphorylation, methylation and ubiquitinylation [39].

The most thoroughly studied epigenetic modification to DNA is methylation. DNA methylation in mammals is almost exclusively restricted to CpG dinucleotides. Clusters of CpG dinucleotides, i.e. CpG islands, are often found within promoter regions of genes where depending on methylation level they either permit or silence transcription. The methylation of CpG dinucleotides can repress transcription either by blocking the binding of transcription factors to the promoter or by recruiting histone-modifying protein complexes that repress transcription through the formation of a more condensed chromatin structure. During cell division, the maintenance methyltransferase 1 is upregulated and recruited to the replication fork where it methylates CpG dinucleotides on the newly synthesized hemi-methylated daughter strand. By the symmetric methylation of the newly synthesized strands, methylation patterns and thus cellular identity is preserved through mitosis [39].

In mammalian development, there are two critical periods of epigenetic modification: gametogenesis and early preimplantation development. During gametogenesis, genome-wide demethylation occurs, which is followed by remethylation before fertilization (see Fig. 1). Early embryogenesis is characterized by a second genome-wide demethylation event, a process that is sensitive to environmental factors [40]. Following implantation, methylation is re-established early in embryonic life.

image

Figure 1.  Methylation reprogramming in the germ line and embryo. The level of methylation in methylated (black) and nonmethylated (grey) imprinted genes and nonimprinted sequences (red, maternal; blue, paternal) is shown during germ-cell and early embryonic development. The horizontal time axis and the vertical axis indicating the relative methylation levels are not to scale. E, embryonic day. Arrows indicate stages in germ-cell and early embryo development in which assisted reproductive techniques might influence and disturb the processes of methylation and/or demethylation. Figure modified from [11].

Download figure to PowerPoint

Genomic imprinting refers to the genetic phenomenon by which certain genes are expressed in a parent-of-origin-specific manner. By definition, only one (maternal or paternal) allele is active, and the inactive allele is epigenetically marked by histone modification, methylation or both [41]. It is currently known that there are more than 80 imprinted genes in mammals. A majority of the imprinted genes have roles in embryonic and placental growth and development. Abnormalities in genomic imprints in humans are known to cause conditions such as BWS and AS (Table 1).

Table 1.   Human phenotypes associated with imprinted genes
SyndromeChromosome location
Beckwith–Wiedemann11p15
Angelman15q11-q12
Prader–Willi15q11-q12
Silver–Russell7p11-p13, 7q31-qter, 11p15
Transient neonatal diabetes mellitus6q24
PHP1b, Albright hereditary osteodystrophy, McCune-Albright20q13
Familial nonchromaffin paraganglioma11q13
Maternal and paternal UPD2, 14, 16
TurnerX

Assisted reproductive techniques involve manipulation of several steps involved in conception that might alter the normal imprinting processes: use of hormones to downregulate pituitary function and to stimulate multiple oocyte production, in vitro maturation of oocytes, use of immature sperm, use of ICSI, in vitro culture of preimplanted embryos and cryopreservation of either gametes or embryos. The stages of germ-cell development and early embryo development in which the ART procedure might influence and disturb the process of methylation and/or demethylation are shown in Fig. 1 (arrows). Hormonal stimulation is frequently used to induce superovulation, and oocytes are retrieved directly from the ovaries prior to ovulation. ARTs also involve in vitro maturation of oocytes prior to fertilization. These steps might influence and disturb the natural process of remethylation during late-stage oocyte maturation (Fig. 1, arrow 1), thereby affecting genomic imprinting of the maturing oocyte. A frequently used method of assisted reproduction is ICSI. Concern has been raised regarding this procedure, because it results in injection of culture medium into the cytoplasm along with the injected sperm (Fig. 1, arrow 2). Furthermore, current ART protocols frequently use prolonged in vitro culturing until the blastocyst stage before embryo transfer, because culture until the blastocyst stage results in higher pregnancy rates. Extended exposure to and/or use of different types of culture medium during early embryo development potentially influences the process of remethylation (Fig. 1, arrow 3).

Animal studies

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Outcome of pregnancies by ARTs
  5. Epigenetics and genomic imprinting
  6. Animal studies
  7. Human studies
  8. Conclusion
  9. Conflict of interest statement
  10. Acknowledgements
  11. References

Animal studies have provided evidence of the effects of ovarian stimulation and embryo culture medium on gene imprinting [4, 12, 42, 43]. Large offspring syndrome (LOS), which is associated with loss of methylation at an imprinting region in the IGF2R gene [44], is a well-known phenotype of cattle born after in vitro culture of embryos [45]. It is characterized by foetal overgrowth, breathing difficulties and sudden perinatal death.

It has been hypothesized that the culture medium causes epigenetic deregulation of developmentally important genes. Such epigenetic alterations would in particular affect the expression of genes that are subjected to genomic imprinting and have key roles in foetal growth control [46]. In a mouse study to determine the effect of culture on imprinted H19 expression and methylation, it was shown that, after culture of two-cell embryos to the blastocyst stage in Whitten’s medium, the normally silent paternal H19 allele was aberrantly expressed, whereas little paternal expression was observed following culture in a different medium (KSOM) containing amino acids (KSOM + AA). Analyses of methylation status at the imprinting control region (ICR) revealed a loss of methylation in embryos cultured in Whitten’s medium but not in embryos cultured in KSOM + AA, indicating that H19 expression and methylation were adversely affected by the culture medium [42].

Another study was conducted in mice to investigate the methylation status of the H19 ICR and H19 promoter with respect to superovulation, IVF and embryo culture conditions. Depending on the culture medium, epigenetic alterations of the H19 ICR and H19 promoter PP were influenced by fertilization, and superovulation clearly disrupted H19 gene expression in individual blastocysts [47]. Recently, the effects of superovulation on genomic imprinting were evaluated in a mouse model, in which effects of inherent subfertility were not a confounding issue. Both maternal and paternal H19 methylation were perturbed by superovulation, in a dose-dependent manner, with changes in imprinted methylation more frequent at higher hormone dosages [48].

Studies have also shown that epigenetic alterations are associated with vitrification, a technique of instant cryopreservation of oocytes or embryos. Wang et al. [49] showed that embryo vitrification aggravated the loss of methylation in the H19/Igf2 DMD (differentially methylated domain, and compensated for the perturbed expression of H19 by IVF procedures). The impact of oocyte vitrification and warming on survival and histone modifications was evaluated in a mouse model, which showed that histone 3 lysine 9 (H3K9) methylation and histone 4 lysine 5 (H4K5) acetylation were increased after oocyte vitrification [50].

Human studies

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Outcome of pregnancies by ARTs
  5. Epigenetics and genomic imprinting
  6. Animal studies
  7. Human studies
  8. Conclusion
  9. Conflict of interest statement
  10. Acknowledgements
  11. References

In view of the results from animal studies, it has become important to establish whether diseases linked to imprinting defects are more prevalent in children conceived through ARTs. Indeed, in 2003, a series of case reports suggested that BWS is more prevalent amongst such children [7, 10, 51]. BWS is a rare disorder with an estimated incidence of 1 in 13 700 infants. As the disease is characterized by overgrowth, it shares many of the features of LOS. BWS is a polygenic disorder and several imprinted loci contribute to disease development, including the Igf2 and H19 loci. However, these initial case reports were criticized for being poorly controlled and for their reliance on questionnaire data to determine the method of conception. These case reports were followed by a case–control study in 2004 by Halliday et al. [52]. Of 37 children with BWS, four were conceived through IVF. In the control group consisting of 148 healthy children, only one child was born after IVF. This study showed that children conceived by IVF are significantly more likely (odds ratio 17.8; 95% confidence interval 1.8–436.9) to have BWS, compared with naturally conceived children. The authors concluded that children with BWS born after IVF procedures had a loss of methylation at the imprinted KvDMR/LIT1 locus. However, in a study of 185 phenotypically normal children, there were no significant differences between those conceived through IVF, through ICSI or naturally at any of the nine differentially methylated regions (DMRs) investigated (including KvDMR1, H19, SNRPN, GRB10, DLK1/MEG3 IG-DMR, GNAS NESP55, GNAS NESPas, GNAS XL-alpha-s and GNAS Exon1A), in the maternal peripheral blood, the umbilical cord blood or the amnion/chorion tissue [53]. Other studies have not identified an increased prevalence of BWS or other imprinting-associated disorders as a consequence of ARTs [27, 29, 33, 54, 55].

Until recently, studies were either small in size, or attempts were made to link the children born with BWS to dysregulated imprinting of genes known to be associated with BWS. Imprinting disorders such as BWS are polygenic, and it is not yet known which additional genomic loci might be influenced through the process of assisted reproduction. Thus, it is possible that the epigenetic effects of these procedures are greatly underestimated. Recent advances in profiling the epigenetics of DNA samples allow for genome-wide approaches to investigate the effects of ARTs on imprinting. This data-driven approach has the potential to identify yet unknown genes that are affected in children conceived through ARTs (i.e. genes that are not necessarily associated with specific clinical manifestations). Recently, Katari et al. [56] investigated patterns of epigenetic modifications using this data-driven approach. The methylation patterns of more than 700 genes were analysed in 10 children conceived in vitro and 13 conceived naturally. The results showed that there were differences in methylation levels of several genes, including some that have been implicated in chronic metabolic disorders, such as obesity and type II diabetes. Furthermore, genome-wide DNA methylation together with chromatin organization was also investigated in human embryos produced by either IVF or ICSI. Similar findings were observed in both groups, leading to the conclusion that ICSI does not result in an increased incidence of epigenetic errors (DNA methylation and chromatin) compared with IVF [57].

Results from studies in sperm suggest that male infertility may contribute to epigenetic effects in pregnancies through ARTs. Altered methylation patterns in imprinted loci have been reported from infertile men (both those with abnormal protamine expression and oligozoospermic patients), compared with fertile donors [58–62]. However, there were differences in the statistical significance of the results between the infertile groups, suggesting that the risk of transmission of epigenetic alterations may differ according to the underlying cause of male infertility. Furthermore, the methylation imprinting marks of two oppositely imprinted genes, H19 and MEST/PEG1, were studied in human testicular spermatozoa from patients with azoospermia of different aetiologies. It was found that spermatozoa from men with abnormal spermatogenesis carry methylation defects in the H19 imprinted gene, which also affect the CCCTC-binding factor-binding site, further supporting an association between the occurrence of imprinting errors and disruptive spermatogenesis [63]. However, it should be noted that these studies do not provide a causal link for epigenetic inheritance of imprinting diseases. More detailed information about the human sperm epigenome can be found elsewhere [58, 60, 64].

Conclusion

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Outcome of pregnancies by ARTs
  5. Epigenetics and genomic imprinting
  6. Animal studies
  7. Human studies
  8. Conclusion
  9. Conflict of interest statement
  10. Acknowledgements
  11. References

Although results from animal studies indicate that ARTs are associated with epigenetic alternations, great caution is recommended in extrapolating these findings to human embryology. One should also keep in mind that the incidence of imprinting disorders is reassuringly low and the great majority of children conceived through ARTs are developing normally.

Acknowledgements

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Outcome of pregnancies by ARTs
  5. Epigenetics and genomic imprinting
  6. Animal studies
  7. Human studies
  8. Conclusion
  9. Conflict of interest statement
  10. Acknowledgements
  11. References

This study was supported by the Seventh Framework Programme of the European Union (FP7/2007–2011) agreement number 259679, the Swedish Research Council (Dnr 2010-2399), the Strategic Research Program in Epidemiology Young Scholar Awards and the Axel and Signe Lagermans Donation Foundation.

References

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Outcome of pregnancies by ARTs
  5. Epigenetics and genomic imprinting
  6. Animal studies
  7. Human studies
  8. Conclusion
  9. Conflict of interest statement
  10. Acknowledgements
  11. References