Abstract
During mammalian development the growth of the fetus is regulated by genetic information that is inherited from both the sperm and the oocyte. Apart from the clear differences that are associated with the X and Y chromosomes, the parental genetic contributions to the embryo also differ via a system of ‘epigenetic’ marks. The differences in function between the parental genomes, how gametes and preimplantation embryos are reprogrammed, and how these delicate processes may be affected by ART and infertility will be described in this chapter. A full understanding of the cellular and molecular biology of human reproduction must include a study of epigenetics and genomic imprinting.
Introduction
During mammalian development the growth of the fetus is regulated by genetic information that is inherited from both the sperm and the oocyte. Apart from the clear differences that are associated with the X and Y chromosomes, the parental genetic contributions to the embryo also differ via a system of ‘epigenetic’ marks. The differences in function between the parental genomes, how gametes and preimplantation embryos are reprogrammed, and how these delicate processes may be affected by ART and infertility will be described in this chapter. A full understanding of the cellular and molecular biology of human reproduction must include a study of epigenetics and genomic imprinting.
Epigenetics
Epigenetics is an additional ‘layer’ of information that complements the information in the genomic DNA sequence; it is essentially a marking system that regulates gene expression and hence the phenotype of the cell. An epigenetic mark either modifies a DNA base sequence chemically, or modifies another molecule (e.g., histone) that leads to a change in chromatin structure. These modifications influence the way that DNA interacts with transcription complexes and other regulatory factors, altering the sequence information that is read by the cell. This level of gene expression is known as the ‘epigenotype’; an incorrect representation of the information, such as occurs in an epigenetic disease, is an ‘epimutation.’ An epigenetic marking system is essential for normal mammalian development, and disruption of the process can lead to disease. Epigenetic marks are extensively reprogrammed during gametogenesis and preimplantation development, subsequently instructing the growth and development of the conceptus. It is therefore crucial to ensure that these intricate processes are adequately supported during ART pathways and manipulations. The biological mechanisms that are regulated via epigenetic modification include X chromosome inactivation and parent-of-origin effects of genomic imprinting, as well as tissue-specific and age-dependent DNA modification. Epigenetic information is also responsible for the phenotypic variability of somatic cells within an organism, controlling how tissues and cells in the body define themselves. For example, within a human individual, the genotype (the gene complement) is the same for different types of cells (e.g., a muscle cell or a liver cell), yet the cellular functions and phenotypes are very different. Although the DNA sequence within these different cells is identical, the repertoire of expressed genes differs greatly between cell types, a difference that is essential in determining the specific functions of different cells.
Epigenetic Marks
The mechanisms that contribute to imprinting include DNA methylation, histone modification and RNA-mediated (transcriptional) mechanisms. DNA methylation is stable but reprogrammable, heritable and affects the regulation of gene expression. This type of epigenetic mark involves methylation of CpG dinucleotides around certain genes. For example, a gene whose active expression is required in a liver cell may be unmethylated across the majority of CpG sites in the promoter area of its sequence, whereas a gene that needs to be silenced in the liver may be heavily methylated, repressing transcription by ‘locking’ the genes within an inaccessible heterochromatin structure.
Histone proteins are important in DNA packaging; they can be covalently modified by a number of post-translational modifications that significantly affect whether chromatin conformation is open or closed. ‘Open’ chromatin is accessible to DNA replication and transcription (gene expression), and ‘closed’ chromatin is not accessible. Chromatin conformation is affected by modification of histone tails via methylation, acetylation, phosphorylation etc., especially those of H3 and H4 histones. Each different modification or combination of modifications affects chromatin structure, and thus gene expression, differently. In many situations, both DNA methylation and the histone modification ‘code’ probably contribute to the overall process of epigenetic regulation.
RNA is increasingly recognized as an important epigenetic regulator, including small noncoding RNA (sncRNA) classes that are expressed in oocytes and preimplantation embryos. For example, microRNAs (miRNAs) are short, noncoding RNA molecules 21 to 24 nucleotides in length that can contribute to an RNA-induced silencing complex; this complex can regulate gene expression by post-transcriptional mechanisms that silence a gene. Surprisingly, mammalian spermatozoa carry sncRNAs: these may be involved in epigenetic processes that are important for the early stages of fertilization (Rivera & Ross, 2013; Yuan et al., 2016).
Histone modifications are named by using:
the name of the histone (e.g., H3)
the abbreviation for the amino acid and its position within the modified histone protein (e.g., K9 for lysine at position 9)
the type of epigenetic modification (e.g., ‘me’ for methylation)
the extent of the particular modification (e.g., mono, di or tri-methylation: me1, me2, me3).
Genomic Imprinting
Genomic imprinting is the exclusive expression of only one of the parental alleles of a gene, a unique mode of gene expression that affects the growth and development of the fetus according to whether an allele of a particular imprinted gene is inherited paternally or maternally. In normal circumstances, imprinting exerts fine control over the growth of the developing conceptus via the placenta. A number of human diseases involve abnormal regulation of imprinted genes or ‘parent of origin’ effects, and disruption of imprinting can lead to cancer (Walter and Paulsen, 2003; Holm et al., 2005). Genomic imprinting is particularly susceptible to disruption during early preimplantation development, and is therefore vulnerable to potential aberrations that may be introduced through certain ART procedures.
Over 200 genes that show imprinted monoallelic expression have been described to date in humans (Morison & Reeve, 1998; Skaar et al., 2012). Genomic imprinting is regulated by imprinting control regions or imprinting centers (ICRs or ICs) that acquire epigenetic marks such as DNA methylation upon passage through the germline. Imprinted genes on maternal and paternal alleles have different methylation patterns (differentially methylated regions, DMRs) that are important in regulating gene expression. DMRs typically consist of stretches of differentially methylated CpG sites that are close to an imprinted gene, and this epigenetic information regulates allele-specific gene expression. A germline DMR will therefore have a different methylation pattern in the sperm than in the oocyte, and this differential marking will be recognized in the zygote and preimplantation embryo. The majority of imprinted genes are located in clusters within the genome and are regulated by ICRs that instruct all or most of the imprinted genes in a cluster.
Figure 15.1 shows a simplified representation of differential allele marking of an imprinted gene in the germline, which leads to monoallelic expression in somatic tissue of the offspring. The paternal allele is not methylated in sperm (left), whilst the maternal allele is methylated in the oocyte (right). This differential methylation imprint persists after fertilization and early development. The epigenetic mark (methylation) placed on the allele during oogenesis silences the maternal allele in the offspring, and only the paternal allele is transcribed to mRNA (bent arrow). The correct dosage of imprinted gene transcripts is critical in early development (Charalambous et al., 2012). Imprinted genes are also believed to play a role in the parent–offspring conflict model (Moore and Haig, 1991), a theory proposing that the parental alleles of these loci have different interests with respect to regulation of fetal, placental and neonatal growth; i.e., the paternal genome will fight for the biggest size and health of the current litter (his litter), whereas the maternal genome will try to counter or moderate this effect in order to reduce current nutritional strain on herself and ensure that she will be able to raise future litters. This hypothesis is reinforced by the fact that many imprinted genes regulate growth of the fetus and placenta. Imprinting may also prevent parthenogenesis, since it ensures that both parental genomes are necessary for an embryo to develop to term (Kono, 2009).
Epigenetic Reprogramming in the Germline
The mature oocyte and sperm are highly specialized cells, and this means that their cellular specialization needs a significant amount of epigenetic information. Information inherited from the previous generation must be erased in primordial germ cells, so that new epigenetic information may be subsequently added according to whether the primordial germ cell is destined to become an oocyte or a sperm cell (Figure 15.2). Extensive epigenetic reprogramming is therefore required in the primordial germ cells (erasure) and during gametogenesis (establishment) (see Morgan et al., 2005 for review). The most complete information available for reprogramming is via DNA methylation, but other epigenetic marks (e.g., the histone code) are also reprogrammed in primordial germ cells and during gametogenesis.
Figure 15.2 The life cycle of genomic imprinting (adapted with permission from Morgan et al., 2005). Imprint erasure, establishment and maintenance are all features of the epigenetic reprogramming cycle that is required for genomic imprinting. In this figure, the imprint is via DNA methylation. Extensive reprogramming occurs in the primordial germ cells, erasing the imprints from the previous generation. Imprints are re-established during gametogenesis according to the sex of the embryo, maintained in the embryo, and translated into stable functional differences between the parental alleles in the developing conceptus. Extensive epigenetic reprogramming of the paternal pronucleus occurs in zygotes.
Erasure
Reprogramming events have been studied extensively in the mouse (reviewed in Constância et al., 1998). Mouse primordial germ cells are identifiable by embryonic Day E7.5, and they then migrate to the genital ridge by Day E10.5–11.5, where they form the gonadal primordia. At this stage, primordial germ cells still retain methylation patterns derived from the oocyte and sperm, and any of their respective modifications from early development. At Day E13.5, male germ cells enter into mitotic arrest and female germ cells enter meiotic prophase. Between Days E10.5 and E13.5 DNA at imprinted genes is globally demethylated in the primordial germ cells of both sexes, as well as at many DNA elements including Intracisternal A Particles, AIPS (retroviral elements containing long terminal repeats), Line 1 sequences (long interspersed nucleotide elements), direct repeats and non-CpG island genes. Each element displays a unique erasure profile, and different degrees of demethylation occur between the various elements (see Lees-Murdock and Walsh, 2008 for review).
Establishment
After methylation has been erased, the germ cells that are still diploid undergo de novo methylation. In mouse gametogenesis, the majority of DNA remethylation occurs around Day E15.5 in both sexes, although timing does vary between the various elements; timing also differs between the male germline and female germline (Lees-Murdock and Walsh, 2008). IAPs, Line 1 sequences and other elements are fully methylated in mature sperm. Certain genes escape this remethylation, for example those containing CpG islands (genomic regions with high densities of CpG nucleotides, which are distinct from the smaller differentially methylated CpG sites of imprinted genes).
Imprinting marks are established according to the sex of the individual (i.e., whether the germ cell is an oogonium or spermatogonium), and most of the imprinted genes acquire methylation in the female germline. The H19 imprinted gene is methylated in the male germline. Despite the fact that epigenetic information has been erased in the primordial germ cells, sufficient underlying epigenetic information remains so that the parental origin of all of the alleles can still be distinguished: for H19 in the male germline, the DMR on the paternal allele is completely remethylated by E15.5, whilst the maternal DMR is remethylated around birth.
In the female germline, maternal DMRs remain hypomethylated until the pachytene stage of meiosis I in the postnatal growing oocyte. Maternal methylation imprints are acquired during oocyte growth, and the DMRs for different imprinted genes appear to be remethylated at different times. Thus, imprinted genes Snrpn, Znf127, and Ndn are methylated by the primary follicle stage, Peg3 and Igf2r genes are methylated by the secondary follicle stage, and the Impact gene is methylated by the antral follicle stage. As in the male germline, some underlying epigenetic signal is retained since the maternally inherited alleles of Snrpn, Zac1 and Peg1 genes are methylated before the paternally inherited alleles (Obata and Kono, 2002; Lucifero et al., 2004; Hiura et al., 2006).
The DNA methyltransferases (Dnmts) play a major role in the establishment of methylation imprints during gametogenesis; the de novo methyltransferases Dnmt3a, Dnmt3b and the related protein Dnmt3L are expressed coordinately and work together to establish methylation imprints during oogenesis. Methylation at imprinted genes in oocytes appears to be dependent on the size of the oocyte, and it has been suggested that methylation is linked to the accumulation of DNA methyltransferases during the growth phase. In oocytes, histone H3K4 must be demethylated (via KDM1B histone demethylase) before the DNA methylation imprints are established; transcription through imprinted gene DMRs keeps chromatin domains open and accessible for methylation (Figure 15.3).
Figure 15.3 Regulation of imprinting in the mouse female germline. The approximate sequence of gene-specific imprint establishment in the mouse female germline is illustrated, showing factors that are relevant to the reprogramming events. Imprinted genes receive methylation imprints at different times during the oocyte growth phase. The lower panel shows key reprogramming phases that occur at imprinted genes and in the genome as a whole. GV = germinal vesicle; MII = metaphase II; PN = pronuclear; ICM = inner cell mass; TE = trophectoderm.
Epigenetic Events during Fertilization and Preimplantation Development
Parental genomes are packaged differently in the gametes and during the first stages of fertilization. The maternal genome from the oocyte is nucleosomal, whilst the paternal genome from the sperm is condensed and packaged mostly by protamines, which are quickly lost and replaced by histones after sperm entry into the oocyte. During pronuclear maturation, the paternal pronucleus undergoes significant chromatin reorganization, with active demethylation during the transition from PN0 to PN5 and metaphase, followed by histone acquisition toward the end of PN maturation. The male pronucleus gathers epigenetic marks such as H3K9me1 and me2, and H3K27me2/3, whilst the maternal pronucleus, largely rich in histone epigenetic marks, remains relatively unchanged from PN0 to PN5. This difference between male and female pronuclei is referred to as ‘epigenetic asymmetry.’ By the end of pronuclear maturation (PN4/PN5), DNA methylation (5MeC) is removed from the male pronucleus. Active demethylation is mediated by the ten-eleven translocation (TET) family of enzymes, which hydroxylate 5-methylcytosine to 5-hydroxymethylcytosine (5-hmC) (Hill et al., 2014; see also Chapter 1, DNA methylation). Differential histone modifications between the parental genomes are observed in the early embryo (reviewed in Corry et al., 2009).
DNA methylation is again reprogrammed during preimplantation development (reviewed in Reik et al., 2001), with initial erasure and then de novo DNA methylation toward the end of the preimplantation period as differentiation occurs. The inner cell mass (ICM) and trophectoderm of the blastocyst show epigenetic differences: H3K27me1, me2 and me3 are present predominantly in the ICM (reviewed in Morgan et al., 2005). The ICM and trophectoderm cell lineages of the mouse blastocyst are differentially marked by histone H3 lysine 27 methylation at key developmental genes.
Genome-wide analysis of methylation can now be carried out via methods based on next-generation sequencing, including whole genome bisulfide sequencing (WGBS) and reduced representation bisulfite sequencing (RRBS). These methods have been used to report comprehensive methylation landscapes in mouse oocytes and preimplantation embryos (Smallwood et al., 2011), and more recently in human gametes and preimplantation embryos (Guo et al., 2014). Results from these novel sequencing-based methods generally support earlier observations based on immunofluorescent methods, i.e., a general decrease in methylation is observed during human preimplantation development. The increased sensitivity of the sequencing-based methods allows the sequences themselves to be identified and has also revealed additional interesting observations during human preimplantation development. Most significantly, despite a decrease in DNA methylation overall, there are two short bursts of remethylation: (i) in the zygote and (ii) at the four- to eight-cell stage. In addition, sensitive single-cell analysis has shown that individual embryonic blastomeres can be traced back to the cell of origin by DNA methylation patterns (Figure 15.4).
Figure 15.4 Assisted reproduction procedures associated with epigenetic errors, based on animal and/or human studies, superimposed on the main events in the epigenetic reprogramming cycle. SECSI = secondary spermatocyte injection; ROSI = round spermatid injection; ROSNI = round spermatid nucleus injection; COH = controlled ovarian hyperstimulation; IVM = in-vitro maturation (of oocytes); GIFT = gamete intrafallopian transfer; ICSI = intracytoplasmic sperm injection; IVF = in-vitro fertilization; IVC = in-vitro culture; ET = embryo transfe; ELSI = elongated sperm injection.
Species-Specific Differences in Mammalian Reprogramming
Assessment of 5-methylcytosine immunostaining in different mammalian zygotes shows:
The paternal pronucleus undergoes demethylation in mouse, human and bovine zygotes, but not in sheep and rabbit zygotes.
Demethylation of the male pronucleus occurs more slowly in rat than in mouse.
The male pronucleus is completely demethylated within 4 hours of fertilization in the mouse, and passive loss of methylation in the embryo continues up to the morula stage.
Imprint Maintenance during Preimplantation Development
Dynamic changes take place in the level of genomic DNA methylation during preimplantation development. The imprints that were established in the male and female germline must be recognized and maintained during global DNA demethylation in the early embryo, so that the imprinting mark may be propagated during later development in order to allow expression of the appropriate allele.
Epigenetic Modification and Assisted Reproductive Technologies
Handling human gametes and embryos outside the human body could potentially introduce stresses that might later be manifested during development. In addition, ART procedures are performed during a period when dynamic and essential epigenetic reprogramming events are occurring on the genome of the gamete or embryo during normal development. Not surprisingly, data surrounding the subject is fragmented and incomplete, fraught with differences in sample size and selection criteria. At the time of writing, an estimated 8 to 10 million healthy children have been born after ART, bearing testimony to this remarkable technology of the twentieth century. However, developmental abnormalities must be rigidly monitored, and research on the epigenetic regulation of human gametes and embryos is fundamental. A summary of some of the syndromes that have been noted in ART children is presented below, together with examples of potential ART pathways that might possibly induce epigenetic mistakes.
Disorders of Genomic Imprinting and Human ART
Several congenital disorders occurring after natural conception that feature disrupted expression of imprinted genes have been recognized as imprinting disorders. Examples of imprinting disorders include Beckwith–Wiedemann syndrome (BWS), Silver–Russell syndrome (SRS) and Angelman syndrome (Odom and Segars, 2010; Eggerman et al., 2015). These disorders can be caused by a variety of mechanisms, including chromosomal causes (e.g., disomy), gene mutations or defects in epigenetic mechanisms. In rare cases, it appears that assisted reproduction may affect epigenetic mechanisms that may result in imprinting disorders. The association between disorders of genomic imprinting and human ART has been extensively summarized (Hiura et al., 2014). A survey of the registries for each of the disorders of genomic imprinting as well as data from case studies of affected individuals were used in order to propose conclusions. A systematic review and meta-analysis of the existing literature on the connection between ART and imprinting disorders published in 2014 indicated that the risk of imprinting disorders is higher in children conceived through ART compared with those who are conceived naturally (Lazaraviciute et al., 2014). Current data indicate that BWS is more prevalent with the use of assisted reproduction. SRS has also been reported as more prevalent with the use of ART. It is important to state that these cases of ART-associated imprinting disorders are rare. However, it is of paramount importance to understand the mechanism and the ART conditions that can cause epigenetic disruption in order to minimize the risk to children born after ART. Earlier publications reported possible associations between the use of ART and Angelman syndrome (AS) and Prader–Willi syndrome (PWS); however, more recent assessments indicate that children with AS and PWS are more likely to be born to parents with a fertility problem, rather than being caused by ART procedures themselves (Vermeiden & Bernardus, 2013).
Beckwith–Wiedemann Syndrome
BWS is caused by faulty expression of the imprinted genes on chromosome 11q15.5 and occurs sporadically after natural conception at an approximate rate of 1 in 15000 births. The syndrome is associated with large pre/postnatal growth (approximately 160% increase), childhood tumors (commonly Wilms’ tumor), macroglossia, exomphalos, organomegaly, hypoglycemia and hemihypertrophy. Approximately 20% of cases show paternal uniparental disomy of the 11q15.5 chromosome, and overexpression of the imprinted gene IGF2 is found in 80% of cases.
BWS registries reveal that the syndrome has been observed in children conceived after assisted reproduction. The major epimutation identified in these children is hypomethylation of the KvDMR1, a methylated imprinting control element on maternal (oocyte) chromosome 11p15.5, at the promoter region of the KCNQ1OT1 gene; 24 out of 25 ART children with BWS presented with hypomethylation at KvDMR1 (Lim et al., 2009). The cause of the BWS epimutation could not be linked to any particular aspect of ART or infertility, but a study that identified 19 ART-conceived children from a BWS registry found the use of ovarian stimulation to be the only common parameter identified (Chang et al., 2005). This aspect will be covered in more detail later in this chapter. Summarizing the data in 2013, Vermeiden and Bernardus described a significant positive association between IVF/ICSI treatment and BWS, with a relative risk of 5.2 (95% confidence interval 1.6–7.4). Mussa et al. (2017) assessed the prevalence of the syndrome in a large group of naturally conceived patients born with BWS in Piemonte, Italy, between 2005 and 2014. A comparison with BWS prevalence in children reported in the corresponding regional ART registry revealed that ART leads to a 10-fold increased risk of BWS compared to the naturally conceived population; however, the report included some BWS cases with nonepigenetic aetiology. Tenorio et al. (2016 ) observed that 88% of BWS patients born after ART had hypomethylation of KvDMR1, compared to 49% for patients with BWS that were conceived naturally.
Silver-Russell Syndrome
SRS is a genetically heterogeneous condition that features growth retardation and learning disabilities. Some cases of SRS are caused by hypomethylation of the IGF2/H19 imprinting center region (ICR1). Several cases of SRS have been described in children conceived by ART, but more studies are required to establish whether there is a true association. Chopra et al. (2010) reported a female child conceived by ICSI who showed hypomethylation of the paternally derived H19/IGF2 locus. The same paper also presented details of six earlier cases of ART-related SRS compiled from other studies. In 2012, a Japanese nationwide epidemiological study reported five cases of ART-related SRS with hypomethylation of the H19/IGF2 locus, but noted that other imprinted loci were also affected (Hiura et al., 2012). Kagami et al. (2007) documented a case of SRS in a child conceived via IVF who showed hypermethylation at the PEG1/MEST DMR, although this particular case had normal methylation at the H19 DMR, the region that is most typically implicated in SRS. Vermeiden and Bernardus (2013) summarized data from several studies, with the conclusion that a significant positive association between SRS and IVF/ICSI treatment is likely.
Angelman Syndrome
AS is a rare disease that affects approximately 1 in 15000 newborns, caused by a spectrum of genetic defects, including a defect in the SNRPN gene on the chromosome 15 imprinting center. An increased incidence of AS has been reported following the use of ICSI (Cox et al., 2002; Ørstavik et al., 2003; Ludwig et al., 2005; Sutcliffe et al., 2006); it has been suggested that some aspects of the ICSI technique might be responsible for the epigenetic abnormality, for example introduction of the sperm acrosome and its digestive enzymes into the ooplasm, ICSI-induced mechanical stress on the oocyte, or disruption of cellular factors or structures required for correct imprinting of chromosome 15. Paternal RNA-mediated mechanisms must also be considered. Loss of methylation at the SNRPN imprinting control region was observed in three cases (normally accounting for less than 5% of all AS cases, occurring in only 1 in 300 000 newborns). Conversely, no SNRPN methylation defects were observed in a study of 92 children born after ICSI (Manning et al., 2000). Sanchez-Albisua et al. (2007) described one AS case born via ICSI with an imprinting defect; Johnson et al. (2018) documented a case of AS with an imprinting defect at SNRPN after IVF. A study by Doornbos et al. (2007) suggests a significant association between fertility problems and AS, but not between fertility treatments (IVF or ICSI) and AS. Vermeiden and Bernardus (2013) also concluded that there is probably a positive association between fertility problems and AS, but no significant association with fertility treatments per se. At the present time, data suggest that AS may be associated with infertility, but are in conflict regarding its association with ART pathways.
Do BWS and SRS Cases Point to Generalized Disruption of Imprinting?
The molecular genetic information from the above examples suggests that a more generalized epigenetic defect may be associated with ART in some cases (e.g., inefficient maintenance of imprints in the preimplantation embryo). In a small number of ART-conceived BWS patients, epigenetic defects outside of KvDMR1 were identified at the DMRs for the imprinted genes IGF2R, SNRPN and PEG1/MEST (Rossignol et al., 2006). However, widespread epigenetic errors are also seen in naturally conceived BWS patients. Hypomethylation at KvDMR1 has also been observed in 3 out of 18 normal children that were conceived by ART (Gomes et al., 2009), supporting the idea that epigenetic defects associated with ART may be more ‘global’ in nature, perhaps insufficient to cause BWS in these cases. Tee et al. (2013) reported that the subgroup of BWS cases with multiple epimutations are preferentially, but not exclusively, conceived with ART.
Epigenetic Changes in Human ART Embryos and ART Cohorts
Epigenetic programming essentially acts as the interface between the environment and the genome, and therefore epigenetic signatures associated with an artificial environment may be evident in preimplantation IVF/ICSI embryos. A legacy of these epigenetic changes may be retained through to adulthood. Despite significant recent advancements, assessing the epigenetic health status of the human ART embryo is not currently practical. With perhaps the exception of the rare cases of imprinting disorders described above, it is not possible to determine with certainty whether any epigenetic changes that may be induced as a result of ART and/or infertility will go on to cause disease in the infant at birth, or lead to long-term developmental consequences. Although our understanding of epigenetic programming in human preimplantation development has improved markedly in recent years, much information is still missing with regard to how this information is regulated during subsequent postimplantation development of the human conceptus. Current data indicate that epigenetic signatures of ART cohorts may differ from naturally conceived cohorts. There have also been reports that human ART embryos themselves may harbor epigenetic changes; however, the major restriction to these studies is that comparison with the epigenetic status of the naturally conceived embryo is not possible. At present we are unable to elucidate whether the epigenetic errors reported in ART cohorts and human ART embryos are the result of the ART procedures, the underlying infertility or a combination of these.
Epigenetic Changes in ART Cohorts
A number of research groups have assessed either gene methylation and/or gene expression differences between cohorts of children born after ART and natural conception (summarized by Batcheller et al., 2011 and Mainigi et al., 2016). Out of 10 studies reported by Mainigi et al. (2016), 7 observed epigenetic differences in ART cohorts. Data as to whether epigenetic signatures of ART cohorts differ from naturally conceived cohorts are currently in conflict. It is important to note that conclusions for some studies will be limited if they were performed using earlier PCR-based DNA methylation analysis of a restricted number of imprinted genes, thought to be the most likely to be susceptible to epigenetic perturbation. More recent studies have used more comprehensive genome-wide assessment of DNA methylation. Using a methylation array-based approach, Melamed et al. (2015) observed hypomethylation and significantly higher variation in DNA methylation in the assisted reproduction group, compared with naturally conceived cohorts. Newly developed methods of methylation analysis such as pyrosequencing allow quantitative assessment of methylation, albeit at a restricted number of loci. Using this method, Whitelaw et al. (2014) indicated that the use of ICSI was associated with a higher level of SNRPN methylation when compared to spontaneous conceptions or IVF offspring. Epigenetic changes induced by ART or infertility may also have detrimental effects on the placenta, although data are again conflicting. Choufani et al. (2018) used genome-wide profiling to reveal methylation loss of several imprinted genes in a subset of ART placentas, and also identified epigenetic profiles in IVF/ICSI placentas that were distinct from those derived from other less invasive ART procedures. Using two different surrogate measures of global DNA methylation (sequence-specific LINE1 assay and Luminometric methylation assay, LUMA), Ghosh et al. (2017) showed that global DNA methylation in IVF placentas differs from that of placentas obtained from natural conceptions. The same study showed that placental DNA methylation was affected by two clinical procedures: oxygen tension and fresh versus frozen embryo transfer. Song et al. (2015) also described differences in placental DNA methylation in children conceived using ART compared to natural conception, and attributed these changes to the ART procedure rather than to a predisposing parental effect. In contrast, Camprubi et al. (2013), using focused methods to assess DNA methylation as well as allelic expression of the imprinted genes, did not identify any defects in placental genomic imprinting after ART.
Epigenetic Changes in ART Embryos
Comprehensive DNA methylation data for human preimplantation embryos has only recently become available. Human studies have been restricted not only by technology, but also because obtaining ART embryos for research is difficult (consent for research and ethical permission). Moreover, human embryos from natural conceptions are not available for use as a control for the investigations that are performed in ART-derived embryos. Studies are also complicated by genetic heterogeneity between human preimplantation embryos. Methylation defects in human ART preimplantation embryos have been identified at DMRs of imprinted genes, but we will never know the naturally occurring rates of these defects in in-vivo conceptions. Geuns et al. (2003) characterized DNA methylation at the imprint control (IC) region of the SNRPN gene in human preimplantation embryos, reporting that the status of a number of embryos appeared to be mainly methylated or mainly unmethylated, possibly indicating an abnormal imprint in these embryos, with a normal imprint in other embryos. White et al. (2015) used traditional bisulfite mutagenesis and sequencing to study imprinted genes in Day 3 and blastocyst stage embryos. Abnormal methylation of imprinted genes/regions SNRPN, KvDMR1/KCNQ1OT1 and H19 was observed at both developmental stages; however, the authors suggest that extended culture to blastocyst stage did not result in embryos with a greater number of epigenetic defects. Khoueiry et al. (2012) reported aberrant methylation of KvDMR1 in abnormal and developmentally delayed embryos that had been fertilized by intracytoplasmic sperm injection (ICSI). In contrast, the same study indicated that the H19 imprint was not affected by embryo grade or developmental delay. Ibala-Romdhane et al. (2011) identified significant hypomethylation as well as hypermethylation of the H19 DMR in arrested preimplantation embryos, whereas methylation of the H19 DMR was normal in embryos that were deemed suitable for transfer. Huntriss et al. (2013) observed contrasting imprinting states for the PEG1/MEST gene in human ART embryos: PEG1/MEST expression was strictly monoallelic in some embryos, and other embryos had bi-allelic PEG1/MEST expression. Therefore, although several studies apparently provide evidence for detection of abnormal methylation and/or expression of imprinted genes in ART-derived embryos, it is important to reiterate the lack of current understanding about the frequency/prevalence of these errors in natural conceptions.
The data described above indicate that it may be interesting to further explore connections between epigenetics and embryo quality/developmental potential. Earlier publications that used immunohistochemical assessment of 5-methyl cytosine also described epigenetic errors in arrested human embryos (Santos et al., 2010). The largest study to date carried out a comprehensive analysis of 57 human blastocysts by WGBS, revealing significant differences in DNA methylation levels between high-quality, middle-quality and low-quality blastocysts; lower quality blastocysts showed more variation in DNA methylation (extremes). The authors also observed an association between the DNA methylome status of the blastocyst and live birth rates (Li et al., 2017).
Epigenetic Changes Attributed to ART Procedures
Elucidating mechanisms that may be detrimental to epigenetic processes is an important goal in order to minimize risk to ART cohorts, with experimental assessment of ART techniques and protocols. Evidence from both human and other mammalian studies suggests that different ART procedures may cause epigenetic changes. Importantly, animal studies are performed in fertile animals, which allows the effect of the ART protocol to be studied in a fertile background with suitable in-vivo–derived controls. In human ART patients, it is difficult to determine whether any detrimental effects are due to the effects of the ART protocols or to infertility itself; furthermore, control embryos are not available from natural conceptions. Current evidence indicates that a number of ART pathways have the potential to cause suboptimal epigenetic programming (see Figure 15.4), but in-depth discussion of all of these areas is beyond the scope of this chapter. ART protocols with potential consequences identified include embryo culture media (formulation, age, addition of growth factors), culture conditions (pH, oxygen tension, build up of waste products), culture period, use of controlled ovarian stimulation/superovulation, embryo transfer, ICSI, embryo biopsy, cryopreservation, in-vitro maturation of oocytes and in-vitro growth. The evidence for epigenetic changes caused by cell culture, controlled ovarian stimulation/superovulation and cryopreservation are discussed in more detail below.
Culture Media and Culture Environment
The preimplantation embryo develops in an artificial environment during ART, and culture media must be suitable and sufficient to support early development, including the capacity to support dynamic epigenetic processes during fertilization and preimplantation development. Our knowledge of how in-vitro culture (IVC) influences the human epigenome is at an early stage. Given the difficulties of using human embryos for research, much of our understanding of how IVC may affect embryo development comes from other mammals, particularly bovine, ovine and mouse data. However, it must be appreciated that not all findings from these studies are immediately relevant to humans, due to species-specific differences in epigenetic programming. A large body of literature has described the effects of IVC on gene expression in preimplantation embryos from several mammalian species (Khosla et al., 2001; Huntriss & Picton, 2008; Denomme & Mann, 2012). Sasaki et al. (1995) showed that IVC mouse embryos experience a loss in H19 imprinting compared to in-vivo-derived embryos. Doherty et al. (2000) and Khosla et al. (2001) concluded that culture conditions can affect the expression of the H19 imprinted gene. Further studies revealed that some types of culture media could alter the expression and/or methylation of a number of imprinted genes in the embryo as well as the placenta (Mann et al., 2004; Market-Velker et al., 2010; Fauque et al., 2007). Earlier studies were restricted by the molecular techniques available at the time, and thus could assess only a small number of genes. More recent data suggest that non-imprinted genes may also be susceptible, indicating that ‘global’ epigenetic changes may be manifested during cell culture. A comprehensive assessment of the effects of culture media was reported by Schwarzer et al. (2012). Mouse preimplantation embryos were exposed to 13 commercially available embryo media used for human ART, and a wide range of cellular and developmental effects on the embryos was assessed. In particular, large-scale assessment of preimplantation embryo gene expression indicated that IVC induced effects on metabolic pathways, suggesting that the embryos might modify their metabolism to accommodate/adapt to the media type used.
It is extremely difficult to perform similar experiments in human preimplantation embryos, and accordingly, only a small number of studies have directly investigated the effects of culture media in human embryos. Kleijkers et al. (2015a) used microarray analysis to compare embryonic gene expression patterns after exposure to either G5 medium or human tubal fluid medium. This study showed that several pathways, including metabolic pathways, were differentially expressed between the two different media. Mantikou et al. (2016) observed differences in the expression of 174 genes when human embryos cultured in G5 medium or HTF medium were compared. This study also reported that the developmental stage of the embryo and maternal age had a greater effect on gene expression than the type of culture media used. Kimber et al. (2008) showed that single growth factors added to human embryo culture media caused unexpected changes in embryonic gene expression profiles.
Numerous animal studies indicate that IVC can potentially cause genome-wide changes in gene regulation. In ruminants, IVC can lead to large offspring syndrome (LOS) in some cases, a condition that causes the fetus to grow excessively large in the womb. LOS offspring have developmental defects, and the large size of the conceptus can cause danger to the mother (Young et al., 1998). Earlier studies of the underlying mechanisms indicated that IVC led to detrimental epigenetic changes at Igf2r (Insulin-like growth factor 2 receptor), and were attributed to the use of serum in the media. More recently, with the benefit of genome-wide transcriptome RNA sequencing analysis, it appears that multiple imprinted loci are affected by IVC in LOS offspring, with loss of imprinting observed across a range of tissues (Chen et al., 2015).
The mechanisms that may cause epigenetic disruption during embryo culture are not fully understood; however, exposure to suboptimal culture media and/or culture environments can lead to alterations in metabolic pathways (Schwarzer et al., 2012; Gad et al., 2012; Kleijkers et al., 2015). It is possible that embryo culture could cause disturbances/deficiencies in 1-carbon metabolic pathways that may affect S-adenosyl methionine-mediated epigenetic regulation during embryonic development (see Chapter 1).
It is important to note that factors in the in-vitro environment other than type of media can affect the epigenetic profile of preimplantation embryos, including oxygen (oxidative stress), temperature, pH and the presence of specific chemicals and additives in culture media and build up of waste products such as ammonia (Lane & Gardner, 1994; Gardner & Kelly, 2017).
The effect of culture media on human birthweight has been the subject of considerable debate, with conflicting results published. Human birthweight data is a useful surrogate for fetal growth, and is used to predict early postnatal growth and long-term risk of cardiometabolic disease. Initially, Dumoulin et al. (2010) observed a significant increase in birthweight after IVC of human embryos in one of two commercially available media tested. Birthweight differences were described in subsequent studies by the same group (Nelissen et al., 2012; Zandstra et al., 2018), and offspring were reported to remain heavier during the first 2 years of life. In contrast, however, a number of other studies found no significant correlation between culture media and birthweight (Lin et al., 2013). Zandstra et al. (2015) described differences in birthweight in 6 out of 11 media comparison studies. In a 2018 follow-up study of 9-year-old children born after IVF/ICSI with two different media, the same authors reported significant differences in body weight, BMI and waist circumference. Reassuringly, no significant differences in cardiovascular development were detected (Zandstra et al., 2018). Several factors that might be implicated in birthweight must be considered, including duration of culture, age of the media and its protein source (Zhu et al., 2014a, 2014b). In contrast, Maas et al. (2016) observed that birthweight was not affected by a number of clinical and laboratory changes over an 18-year period, reporting no differences in birthweight when variables such as different media, laboratory location, use of gonadotrophins, IVF or ICSI, and day of transfer were compared. A significant difference in birthweight was observed only between fresh and frozen transfers. Clearly, more research is required to determine whether different media formulations can affect birthweight, subsequent development and long-term health. In 2016, an ESHRE working party called for tracking of culture media use in ART registries with long-term assessment of health risks, together with a call for disclosure of media composition by commercial manufacturers (Sunde et al., 2016). These and other recommendations were echoed by others (Ménézo et al., 2018; Huntriss et al., 2018).
Controlled Ovarian Hyperstimulation (COH)/Superovulation
Animal and human studies both suggest that superovulation can cause epigenetic errors in the oocyte, the embryo and the placenta (Fauque et al., 2013). Ovarian stimulation may drive oocyte maturation within an inappropriate time frame, or in a cellular and developmental context that is incompatible with achieving complete epigenetic programming of the oocyte. It is possible that ovarian stimulation may override the progressive processes of epigenetic maturation and imprint establishment that are connected with oocyte growth and size (Obata and Kono, 2002; O’Doherty et al., 2012). Superovulation could interfere with this stepwise process, recruiting young follicles that have not correctly established their imprinting during maturation. Alternatively, ovarian stimulation may lead to the recruitment of poor quality oocytes that would not be selected to ovulate under normal circumstances, pushing lower quality oocytes to maturity (Market-Velker et al., 2010; Van der Auwera and D’Hooghe, 2001). Due to the difficulties intrinsic to determining the epigenetic status of human oocytes, there are few reports to date; however, controlled ovarian hyperstimulation has been associated with epigenetic changes at a small number of loci (Sato et al., 2007; Khoueiry et al., 2008). As described above, COH was found to be the common treatment in 19 children with BWS that had been conceived through ART (Chang et al., 2005).
Mouse studies have been particularly important in assessing effects of superovulation, but these findings might not necessarily translate between the differing reproductive physiologies of mice and humans. Ovarian stimulation may have transgenerational effects, i.e., induced epigenetic changes can persist in the sperm of second generation male offspring that are born from superovulated female mice (Stouder et al., 2009). Mouse experiments have shown that superovulation can cause aberrant genomic imprinting of both maternally and paternally expressed genes in the embryo and placenta (Fortier et al., 2008; Market-Velker et al., 2010). Disruption of paternally imprinted genes in addition to maternally imprinted genes indicates that ovarian stimulation has the capacity to disrupt key epigenetic ‘master’ regulators, or perhaps other epigenetic processes or related oocyte/early embryonic structures that are required for maintenance of genomic imprinting during preimplantation development (Nakamura et al., 2007; Denomme et al., 2011; Huffman et al., 2015). The effects of superovulation assessed at later time points after implantation indicated that superovulation can affect oocytes sufficiently to cause abnormal imprinted gene expression (Igf2) in the placenta (Fortier et al., 2014).
Cryopreservation
Experimental data suggest that cryopreservation can alter epigenetic marks in mammalian cells (Kobayashi et al., 2009), although the mechanisms causing this damage are unclear at present. These could include general cell damage and/or damage to structures associated with epigenetic programming in oocytes/embryos, or perhaps damage to the spindle. As with all studies of this nature, reports of potentially detrimental consequences are in conflict. Table 15.1 shows a summary of epigenetic consequences observed in several species after vitrification.
Table 15.1 Epigenetic consequences observed after vitrification
| Species | Method | Cells | Mark | Method | Difference? | Reference |
|---|---|---|---|---|---|---|
| Human | Oocyte vitrification | Day 3 embryos | 5mC/5hmc | Immunofluorescence | No difference | De Munck et al., 2015 |
| Mouse | Oocyte vitrification | 5mC | Pyrosequencing (Snrpn, Igf2r) | No difference in imprinted DMR but differences at sporadic CpGs | Trapphoff et al., 2010 | |
| (Preantral follicle) | ||||||
| Mouse | Oocyte vitrification | Oocytes and embryos | 5mC | Immunofluorescence | Lower global DNA methylation | Liang et al., 2014 |
| Rabbit | Embryo vitrification | Embryo | 5mC | Bisulfite sequencing | No difference | Saenz-de Juano et al., 2014 |
| OCT4 promoter | ||||||
| Mouse | Oocyte vitrification | Oocytes and embryos | 5mC | Bisulfite sequencing | Lower DNA methylation | Cheng et al., 2014 |
| H19, Peg3, Snrpn | +Dnmt expression lower | |||||
| Mouse | Embryo vitrification | Embryo | 5mC | Bisulfite sequencing Grb10 | Lower Grb10 DNA methylation | Yao et al., 2017 |
| Immunofluorescence | Lower Grb10 expression | |||||
| Lower global DNA methylation | ||||||
| Mouse | Oocyte vitrification | Oocytes | 5mC | Bisulfite sequencing Dnmt1o, Hdac1, Hat1 | No difference in DNA methylation | Zhao et al., 2013 |
| but | ||||||
| Decreased Dnmt1o expression | ||||||
| Bovine | Oocyte vitrification | Oocyte | 5mC and histone marks | Bisulfite sequencing Immunofluorescence | Difference in one epigenetic mark in TE only | Chen H et al., 2016 |
| Difference in imprinted gene expression in blastocysts | ||||||
| Mouse | Oocyte vitrification | Oocyte | Histone marks | Immunofluorescence | H3K9 methylation increased | Yan et al., 2010 |
| H4K5 acetylation increased | ||||||
| Mouse | Embryo vitrification | Fetus and placenta | 5mC | Bisulfite sequencing H19/Igf2 | Loss of H19 methylation | Wang et al., 2010 |
| Altered H19 expression | ||||||
| Human | Embryo vitrification | Embryo | 5mC | Bisulfite sequencing H19/Igf2 | No difference | Derakhshan-Horeh et al., 2016 |
Infertility and Epigenetics
Chapter 3 outlines the fact that gametogenesis is a complex process, requiring coordination of numerous cellular and molecular events over extended periods of time. Some cases of infertility may be due to suboptimal gametogenesis, and there may be disruption of important epigenetic processes, causing the gamete itself to carry epigenetic defects. Further research is required to fully understand the impact of any gamete-borne epigenetic changes on the process of fertilization and subsequent embryo development – an area that remains a subject of debate. For example, some studies have shown that epigenetic errors may be inherited from the sperm (Kobayashi et al., 2009). However, other studies suggest that epigenetic defects are due to the assisted reproduction procedures rather than defects present in the gametes (Song et al., 2015). It is possible that epigenetic defects present in the gametes could be exacerbated by suboptimal conditions in assisted reproduction.
The cause of these epigenetic defects in the infertile patient may be a result of general defects in gene expression patterns impacting on epigenetic processes, a legacy of the defective process of gametogenesis. A small number of cases could be caused by a genetic mutation in the gamete’s epigenetic machinery. For example, an association has been found between DNA methylation defects observed at imprinted loci in ART concepti and mutations in the gene encoding DNMT3L, a gene that is important in imprint establishment. These mutations were also present in parental sperm (Kobayashi et al., 2009). However, many factors can influence epigenetic programming in the mammalian germline, some of which will influence the physiology of the offspring, including diet and dietary supplements, body composition, advanced age and use of medicines. Environmental exposures and lifestyle factors such as smoking are also important. Finally, genetic/epigenetic variation will also influence gametic programming (Rajender et al., 2011; Boissonnas et al., 2013; Ge et al., 2014; Jenkins et al., 2014; Soubry et al., 2014; Shea et al., 2015; Stuppia et al., 2015; de Castro et al., 2016).
Epigenetic Signatures of Infertility
DNA Methylation
Defective epigenetic signatures associated with a number of different classes of human male infertility are now well documented. Equivalent epigenetic defects in the female germline may be associated with female infertility, but oocytes are far less amenable to epigenetic research than sperm. Many research groups have compared DNA methylation marks at a number of key genes in sperm from infertile males with those from fertile controls, with significant focus on imprinted genes, revealing perturbed sperm DNA methylation signatures at imprinted genes as well as other sequences in cases of male infertility. A meta-analysis including 24 of these studies concluded that male infertility is associated with altered sperm methylation at three imprinted genes (H19, SNRPN, MEST) (Santi et al., 2017). Although imprinted genes are a particularly important group of genes to study, it is likely that infertility can lead to genome-wide changes in DNA methylation compared to that observed in the typical fertile gamete. Modern methods of epigenetic screening, particularly DNA methylation arrays or analysis based on next-generation sequencing, now facilitate understanding this bigger picture, allowing epigenetic marks present in mature, functional gametes and those affected in infertile gametes to be recognized. Global DNA methylation analysis methods indicate that poor quality human sperm may be due to defects in the process of DNA methylation erasure (Houshdaran et al., 2007). Genome-wide DNA methylation analysis by sequencing and array methods has also been used to identify potential epigenetic markers of male infertility for inclusion in infertility screening panels (Sujit et al., 2018).
Protamines
The epigenetic status of a gamete is dictated by systems other than DNA methylation. During human spermiogenesis, the transition from histones to protamines is an important step. The ratio of Protamine 1 to Protamine 2 (P1/P2) is normally 0.8–1.2, and this ratio appears to be important in fertility; disruption of this ratio may be associated with decreased embryo quality and poor IVF outcomes (Aoki et al., 2005, 2006a, 2006b). Abnormal P1/P2 ratios and low protamine levels are associated with increased DNA fragmentation, suggesting that incorrect compaction exposes DNA to oxidative stress and damage (Aoki et al., 2006b; Torregrosa et al., 2006). Histones are not completely replaced by protamines, being retained at a small number of places in the sperm genome, e.g., at key regulatory regions of the genome kept ‘poised’ for activation in early embryonic development. These marks may be major determinants of embryo developmental outcome (Denomme et al., 2017).
Histone Modification
Histone modification plays a crucial role in spermatogenesis. Histone tails are subjected to post-translational modification (methylation, acetylation, phosphorylation and ubiquitination) on different amino acid residues. Histone acetyl transferases (HATs) acetylate lysine residues (K), particularly on histones H3 and H4. This has the effect of relaxing chromatin and making it accessible to transcription factors, and histone acetylation is thus generally associated with gene activation. In contrast, deacetylation by histone deacetylases (HDACs) often leads to gene silencing. The disruption of histone acetylation in spermatogenesis can lead to severe male infertility (Fenic et al., 2004, 2008; Ge et al., 2014). Histone methylation patterns are dynamic during spermatogenesis; the timing of their establishment and removal is critical, controlled by a number of specific enzymes, including histone methyltransferases (HMTs) and histone demethylases (HDM). Generally, methylation of H3K4 is associated with gene expression whilst methylation of H3K9 and H3K27 is linked to gene silencing (Okada et al., 2007; Sikienka et al., 2015).
RNA-Mediated Epigenetic Processes As a Major Epigenetic Determinant of Fertility
RNAs play an important role in germline epigenetic regulation. Human sperm RNA profiles have been shown to differ between fertile and infertile men (Ostermeier et al., 2002). Surprisingly, the sperm nucleus contains many RNAs – not only messenger RNAs (mRNAs), but also sncRNAs, such as miRNAs, piRNAs and sperm tRNA-derived small RNAs (tsRNAs) (Chen Q et al., 2016; Sharma et al., 2016). Some of these RNA species may be particularly important during fertilization (Gross et al., 2017). RNAs play an important role in the epigenetic control of retrotransposons in the germline. Approximately 45% of the human genome is derived from transposable elements, the majority of which originate from retrotransposons, and these can drive their own genomic replication. Insertional mutagenesis results in uncontrolled retrotransposon activity, causing genomic instability and cell death – this must be suppressed in the germline. Retrotransposons are suppressed by DNA methylation in somatic cells. However, DNA methylation is being reprogrammed and is temporarily lost from most of the sperm genome during germline epigenetic reprogramming, and a different mechanism for retrotransposon suppression is employed. piRNAs, together with other factors, facilitate DNA methylation at these elements (Aravin & Hannon, 2008; Frost et al., 2010; Pastor et al., 2014).
Looking to the Future: Epigenetics and the ART Laboratory
Severe technical limitations have hampered epigenetic studies of human preimplantation embryos to date, limiting investigations to a restricted number of loci. These have recently been largely overcome, and refinement of these and other techniques that facilitate work on single embryos, oocytes and even single cells has considerably expanded our ability to understand epigenetic regulation in early human development. Mapping the origins and mechanisms that underlie ART-induced epigenetic defects on a patient-specific basis is now on the horizon.
Sampling of embryonic DNA methylation (for example through trophectoderm biopsy and single-cell sensitive WGBS) remains of debatable predictive value in human ART without further research; however, the technologies to enable this at least for DNA methylation analysis are effectively in place.
The most immediate and practical use of epigenetic screening methods may be their application in the diagnosis of male infertility. A number of groups advocate diagnostic epigenetic screening of sperm for cases of male infertility (Aston & Carrell, 2014; Hotaling & Carrell, 2014; Klaver & Gromoll, 2014). This approach is likely to require extensive validation, as well as further research on human developmental epigenetics.
Research goals for the future include assessing epigenetic marks associated with low pregnancy rates, poor embryo development, potential risks of disease transfer to the offspring and better categorization of infertility aetiologies in order to optimize the treatment path.
Summary
The human embryo is exquisitely sensitive to changes in its environment; ART protocols have been evolving and changing rapidly over the past 40 years, and the fact that so many healthy children have been born despite numerous differences in in-vitro embryo culture is a testament to human developmental plasticity. This biological phenomenon is crucial in ART practice: although it is clear that techniques and protocols can have an effect at the molecular and cellular level, optimal conditions are not yet defined, and the extent of effects on health in the short or long term are unclear (Roseboom, 2018). An association between ART and an increased risk of at least some epigenetic disorders must clearly be considered (Figure 15.5), and sophisticated molecular studies that compare the ‘global’ epigenetic status of children born in vivo and those born in vitro are now emerging. Studies such as these will help us to understand the molecular processes that are affected by ART, and the potential risks associated with each technique. Research techniques have now been developed that allow the epigenetic effects of ART technologies to be rigorously tested to precisely gauge their effect upon the epigenetic development of human gametes and preimplantation embryos so that ART protocols can be adapted to avoid potential problems. Animal studies are useful in highlighting problems associated with particular ART methods, but there are significant differences between humans and other mammals with respect to epigenetic regulation, and in the regulation of imprinting during gametogenesis and early development. Further research on the epigenetic regulation of human gametes and preimplantation embryos is urgently needed in order to understand how infertility, as well as lifestyle factors, impact the epigenome of the gametes, and whether ART can exacerbate these errors. A focus on safety aspects of existing and emerging ART treatments is essential, together with long-term follow-up studies of children born as a result of ART.
Figure 15.5 Major DNA methylation reprogramming events during mammalian gametogenesis and preimplantation development shown together with a summary of the assisted reproductive technology (ART) procedures that are associated with epigenetic errors. The methylation reprogramming panel (center panel, adapted from Reik et al., 2001) indicates the level of methylation in male (M) and female (F) gametes (left side), and also in the paternally inherited (M) and the maternally inherited genomes (F) after fertilization, during preimplantation development (right side). The timing and nature of these reprogramming events varies between species. The dashed lines indicate the maintenance of differential methylation at imprinted genes during preimplantation development. The genome is remethylated differentially within the blastocyst in the embryonic (EM) and extraembryonic lineages (EX).
Further Reading
Reviews
Publications
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
