Epidemiological studies have reported a higher incidence of rare disorders involving imprinted genes among children conceived using assisted reproductive technology (ART), suggesting that ART procedures may be disruptive to imprinted gene methylation patterns. We examined intra- and inter-individual variation in DNA methylation at the differentially methylated regions (DMRs) of the IGF2/H19 and IGF2R loci in a population of children conceived in vitro or in vivo. We found substantial variation in allele-specific methylation at both loci in both groups. Aberrant methylation of the maternal IGF2/H19 DMR was more common in the in vitro group, and the overall variance was also significantly greater in the in vitro group. We estimated the number of trophoblast stem cells in each group based on approximation of the variance of the binomial distribution of IGF2/H19 methylation ratios, as well as the distribution of X chromosome inactivation scores in placenta. Both of these independent measures indicated that placentas of the in vitro group were derived from fewer stem cells than the in vivo conceived group. Both IGF2 and H19 mRNAs were significantly lower in placenta from the in vitro group. Although average birth weight was lower in the in vitro group, we found no correlation between birth weight and IGF2 or IGF2R transcript levels or the ratio of IGF2/IGF2R transcript levels. Our results show that in vitro conception is associated with aberrant methylation patterns at the IGF2/H19 locus. However, very little of the inter- or intra-individual variation in H19 or IGF2 mRNA levels can be explained by differences in maternal DMR DNA methylation, in contrast to the expectations of current transcriptional imprinting models. Extraembryonic tissues of embryos cultured in vitro appear to be derived from fewer trophoblast stem cells. It is possible that this developmental difference has an effect on placental and fetal growth.
We have screened a population of children conceived in vitro for epigenetic alterations at two loci that carry parent-of-origin specific methylation marks. We made the observation that epigenetic variability was greater in extraembryonic tissues than embryonic tissues in both groups, as has also been demonstrated in the mouse. The greater level of intra-individual variation in extraembryonic tissues of the in vitro group appears to result from these embryos having fewer trophoblast stem cells. We also made the unexpected observation that variability in parental origin-dependent epigenetic marking was poorly correlated with gene expression. In fact, there is such a high level of inter-individual variation in IGF2 transcript level that the presumed half-fold reduction in IGF2 mRNA accounted for by proper transcriptional imprinting versus complete loss of imprinting would account for less than 5% of the total population variance. Given this level of variability in the expression of an imprinted gene, the presumed operation of “parental conflict” as the selective force acting to maintain imprinted gene expression at the IGF2/H19 locus in the human should be revisited.
Citation: Turan N, Katari S, Gerson LF, Chalian R, Foster MW, Gaughan JP, et al. (2010) Inter- and Intra-Individual Variation in Allele-Specific DNA Methylation and Gene Expression in Children Conceived using Assisted Reproductive Technology. PLoS Genet 6(7): e1001033. https://doi.org/10.1371/journal.pgen.1001033
Editor: Jeannie T. Lee, Massachusetts General Hospital, Howard Hughes Medical Institute, United States of America
Received: December 18, 2009; Accepted: June 17, 2010; Published: July 22, 2010
Copyright: © 2010 Turan et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the National Institutes of Health (R01 HD048730 to CS and CC). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Several epidemiological studies have reported a higher incidence of rare disorders involving imprinted genes (Angelman syndrome – and Beckwith-Wiedemann syndrome –) among children conceived using assisted reproductive technologies (ART). Studies on imprinted gene expression and parental allele-specific DNA methylation in animal models have also suggested that epigenetic marks may be altered by treatments and procedures commonly employed in ART (ovarian stimulation, egg retrieval, in vitro fertilization, intracytoplasmic sperm injection, preimplantation embryo culture, embryo transfer) –.
CpG sites in differentially methylated regions (DMRs) of imprinted genes are methylated on the allele contributed by one parent and unmethylated on the allele contributed by the other. This pattern of differential allelic methylation is established during male and female gametogenesis ,  and the differences are maintained after fertilization such that cells from most somatic tissues are expected to exhibit the same parental allele-specific methylation pattern , . However, if ART treatments and procedures in the human are disruptive to imprinted gene methylation patterns, as they are in the mouse , one might predict that alterations could occur in some cells of the early embryo but not in others, resulting in individuals who were mosaic to varying degrees for loss or relaxation of proper imprinted allelic methylation. In addition, different degrees of relaxation of allele-specific methylation could be observed between different tissues of the same individual, depending on when the disruption occurred during development.
One of the loci shown to be susceptible to alteration of epigenetic modifications by in vitro culture and ovarian stimulation in the mouse is Igf2/H19 , , –. Because IGF2 is an important placental growth factor and one of the phenotypes most strongly associated with human ART procedures is low birth weight, we reasoned that the human IGF2/H19 locus might also be susceptible during ART treatments and procedures. We compared parental allele-specific methylation between children conceived in vitro or in vivo at the DMR that functions as an imprint control region (ICR) at IGF2/H19 ,  and also at an IGF2 receptor (IGF2R) DMR , . We examined a sample of cord blood, a section of umbilical cord and five sections of placenta in each child for abnormal methylation of maternal alleles at the IGF2/H19 ICR/DMR and for abnormal methylation of paternal alleles at the IGF2R DMR . Under our null hypothesis, little variation in parental allele-specific methylation was expected within an individual, or between individuals, because the methylation status of each CpG site in the DMR is set in the gametes  and faithful replication of this status during development is expected to result in the same allelic methylation ratio in each individual and in each tissue. We also measured steady-state IGF2, H19 and IGF2R transcript levels to determine whether mRNA levels were correlated with abnormal allelic methylation ratio or birth weight. To our knowledge, this is the first study to examine intra-individual variation in epigenetic markings in children conceived using assisted reproduction.
Intra- and Inter-Individual Variation in DNA Methylation at the IGF2/H19 DMR
We investigated intra- and inter-individual variation in allele-specific methylation at the IGF2/H19 DMR. We measured the relative level of CpG methylation on maternal and paternal alleles at this locus in cord blood, cord and five sections of placenta taken from a population of children conceived either in vitro or in vivo.
The imprinted IGF2/H19 DMR is located between the IGF2 and H19 genes and is normally methylated on only the paternal allele , . We used a single nucleotide polymorphism (a C/T SNP at a CfoI site) to identify informative (heterozygous) individuals and a methylation-sensitive restriction endonuclease (MluI) to determine the methylation status of a specific CpG site within the DMR, as described previously . Methylation at the MluI site, and an adjacent MaeII site, have been shown previously to be characteristic of the methylation at surrounding CpGs by bisulfite sequencing (Fig. 7 in Sandovici et al., 2003) . We identified 45 in vitro and 56 in vivo individuals who were informative and for whom DNA was available from cord blood, cord and five sections of placenta.
Because previous studies have indicated that loss or relaxation of imprinting is a quantitative trait , , we measured the ratio between DNA methylation levels on maternal and paternal (M/P) alleles as an indicator of imprinting status. A ratio of zero corresponds to exclusive methylation of the MluI site on the paternal allele, while a ratio of one signifies methylation of this site on an equal number of maternal and paternal alleles (n.b.: Although we have not determined the parental origin of each allele in the present study because DNA was not available from parents, we have shown previously that the less methylated allele was maternal in all 163 individuals for whom we were able to determine parental origin by pedigree analysis . We therefore assume that the less methylated allele is maternal in the population examined here, also.) In the case of controls, no uncleaved C alleles were detected in any C/C homozygous individuals, indicating that CfoI cleaved the “hot-stop” PCR products with >99% efficiency .
Figure 1 shows the distribution of M/P methylation ratios observed from 56 informative in vivo (Figure 1A) and 45 informative in vitro (Figure 1B) individuals. The M/P ratios were measured for cord blood, cord and five sections of placenta from each informative individual. The data are represented as a series of symbols on a vertical line, ranked from individuals showing the greatest range of variation on the left side of the graph to individuals showing the least range of variation on the right side of the graph. The distribution of individual M/P methylation ratios in cord blood (red circles in Figure 1) shows that the great majority of informative individuals have <15% methylation on the maternal allele in both groups. Approximately 8% of the population examined here has ≥15% of methylation of CpG sites on the maternal allele in cord blood. These findings are similar to those reported previously by Sandovici et al. (2003) for M/P methylation ratios measured in peripheral blood samples from the CEPH families . The distribution of individual M/P methylation ratios in cord showed a similar pattern to that observed in cord blood (Figure 1), while the five sections of placenta taken from each individual showed a broad range of intra-individual variation in M/P methylation ratios.
(A) 56 informative in vivo individuals, and (B) 45 informative in vitro individuals. M/P represents the DNA methylation ratio between maternal and paternal alleles (see text). Each individual is represented as vertical data-set of M/P ratios from cord blood, cord and five sections of placenta. Individuals were ranked based on the degree of scatter (intra-individual variation) observed.
When individual tissues are compared, both M/P ratio mean and variance are greater in the in vitro group in each tissue, although two of these comparisons (cord blood means, in which the fewest samples are compared between the two groups, and placenta variance, in which the greatest range of variation is observed) do not reach statistical significance (Table 1). However, because the a priori expectation for the mean M/P methylation ratio of the IGF2/H19 DMR is near zero, independent of the embryonic origin of the tissue (because the methylation status of maternal and paternal DMR alleles is assumed to be determined in the gametes and to escape the genome wide demethylation/remethylation that occurs in preimplantation embryos) , it is not inappropriate to combine all samples in each group to determine whether the two groups differ in mean M/P ratio and whether the groups have equal variance. When samples from all tissues are combined, both allele-specific methylation ratio mean and variance are significantly greater in the in vitro group (P = 0.0001 and P = 0.0006, respectively, Table 1). The fact that the intra-individual variation in M/P ratios is also greater in the in vitro group may be seen, simply, by comparing the fraction of individuals in each group in which all samples have M/P ratios below any arbitrarily chosen value. For example, only 31% of in vitro individuals (14/45) maintain M/P ratios of <0.1 in cord blood, cord and all five sections of placenta while 46% of in vivo individuals (26/56) are below this threshold.
Greater Intra-Individual Variance in Placenta M/P Ratios Is Associated with Fewer Trophoblast Stem Cells
One mechanism by which greater variance in an epigenetic character may occur is through a sampling effect that depends on the number of stem cell progenitors that give rise to a particular tissue; the fewer the number of stem cells, the greater the variance. Because much of the difference in variance observed between the two groups occurs as a result of intra-individual differences in umbilical cord and placenta samples, we estimated the number of trophoblast stem cells that give rise to the placenta in each group by comparing the distribution of X-inactivation scores in females from each comparison group ,  and by comparing the distribution of M/P IGF2/H19 methylation ratios from Figure 1. We note that the assay used in each case amounts to a simple yes/no binomial trial of the form “is the CpG site being examined methylated (in which case it gives a signal) or not (in which case it does not)”, each of which is expected to yield a “success” (the DNA molecule in question is methylated) with probability “p” (p = 0.5 in the case of which X chromosome is inactivated and p = 0.1 in the case of methylation of the maternal IGF2/H19 DMR, see below) or a failure with probability “q” (which is equal to 1-p).
The number of trophoblast stem cells may be estimated from the distribution of X-inactivation scores by comparing the actual distribution of X-inactivation scores with the distribution estimated from the variance of the binomial distribution (pq/N), setting the probability that either allele is methylated (p or q) to 0.5 and generating the distribution for different values of N (number of stem cells) , . We determined the X-inactivation score distribution for each group (using DNA samples from five sections of placenta from 50 in vitro and 54 in vivo females) by comparing allele-specific methylation at HpaII sites adjacent to the CAG trinucleotide repeat in the highly polymorphic Androgen Receptor locus (AR) . The closest fit to the distribution of X-inactivation scores in placenta in children conceived in vitro corresponds to nine trophoblast stem cells and the closest fit in children conceived in vivo corresponds to 11 trophoblast stem cells (Table 2).
We additionally estimated the number of trophoblast stem cells in each group by comparing the distribution of M/P IGF2/H19 DMR methylation ratios (Figure 1) from the two comparison groups, using 0.1 and 0.9 as values for p and q (these values were selected based on the observation that ∼10% of individuals have significant methylation on the maternal DMR while ∼90% have very few cells carrying maternal DMR methylation . These values are also in close agreement with the probability that any maternal DMR DNA molecule is methylated (placenta in vivo mean = 0.0801, in vitro mean = 0.1017, Table 1). Using these parameters, the closest fit to the distribution of M/P ratios in placenta in children conceived in vitro corresponds to eight trophoblast stem cells and the closest fit in children conceived in vivo corresponds to 10 trophoblast stem cells (Table 2).
Overall, these very similar independent estimates of between-group epigenetic variation (n.b.: not only are the two loci examined on different chromosomes but many of the individuals in the X-inactivation groups, composed of only females, and the IGF2/H19 groups were different) are consistent with the prediction that overall greater variance in M/P IGF2/H19 DMR methylation ratios in the in vitro group is associated with fewer trophoblast stem cells.
Analyses of IGF2/H19 DMR Methylation in Placenta by Pyrosequencing
Because the methylation-sensitive restriction endonuclease assay used to generate the data shown in Figure 1 provides a ratio of maternal alleles at which MluI sites are methylated to paternal alleles at which MluI sites are methylated rather than an absolute fraction of all alleles, we also assayed DNA methylation at the IGF2/H19 DMR by bisulfite pyrosequencing. The assay we used quantifies the methylation status of five CpGs in the IGF2/H19 DMR that are adjacent to the CpG queried in the MluI assay. If an M/P ratio greater than zero (Figure 1) represents methylation of maternal alleles in addition to methylation of DMRs on all paternal alleles, then the CpG sites in these samples/individuals should be methylated on greater than 50% of molecules assayed by bisulfite pyrosequencing (i.e. all paternal alleles plus some fraction of maternal alleles). The placenta samples with M/P ratios greater than zero show greater than 50% methylation at all five CpGs in almost all cases (Figure 2) by bisulfite pyrosequencing, indicating that both the methylation-sensitive restriction endonuclease assay and the bisulfite pyrosequencing assay are measuring gain of methylation on maternal alleles, as has also been reported in the mouse , , .
(A) 31 in vivo individuals, and (B) 26 in vitro individuals. The methylation results at each CpG site for each individual are represented by a horizontal series of symbols. All but a few of the CpGs show methylation levels of 50% or higher, which is in concordance with our findings of gain of methylation on the maternal allele (M/P ≥0.15 in placenta, Figure 1).
Allele-Specific Expression of IGF2/H19 Is Not Disrupted by Bi-Allelic Methylation of the DMR
The current model for imprinted transcriptional control of IGF2/H19 correlates methylation at the ICR/DMR with an inability to bind CTCF, transcriptional silencing of H19 and transcriptional activation of IGF2 , . Because our analysis of DMR methylation indicates that all paternal and some maternal DMRs are methylated, we examined mRNA expression in individuals with varying levels of bi-allelic DMR methylation, for the presence of bi-allelic transcripts of IGF2 and H19. We assayed individuals who were informative for an ApaI polymorphism (in exon 9) for the presence of transcripts from both IGF2 alleles . We also assayed individuals who were informative for an RsaI polymorphism (in exon 5) for the presence of transcripts from both H19 alleles , which would be expected to occur if paternal DMRs became demethylated.
Although we detected minor amounts of presumed maternal IGF2 mRNA in several individuals, there was no correlation between maternal DMR methylation and amount of transcript from the maternal allele (Table 3). We also observed only small amounts of presumed paternal allele expression of H19 mRNA, which suggested that there was no loss of methylation on paternal alleles in these samples (Table 3), also consistent with the overall greater than 50% methylation observed by pyrosequencing (Figure 2).
Intra-Individual Variation in DNA Methylation at the IGF2R DMR
We investigated intra-individual variation in allele-specific methylation at an IGF2R DMR we have also examined previously . The rationale for examining allele-specific methylation at this non-transcriptionally imprinted locus is that this locus may be a more sensitive reporter of any disruption in CpG site methylation by environmental factors because such changes are not predicted to affect transcription and are less likely to be selected against. The relative level of CpG methylation on paternal and maternal alleles at this locus was measured in cord blood, cord and five sections of placenta taken from populations of children conceived either in vitro or in vivo. The less methylated allele is assumed to be paternal because we found no individuals in which the paternal allele was more methylated than the maternal allele in 112 informative individuals for whom allelic inheritance could be confirmed by pedigree analysis .
The parental origin-specifically methylated IGF2R DMR is located in the second intron of IGF2R and is normally methylated on the maternal allele. Although the human IGF2R gene is transcribed from both alleles  differential methylation of maternal and paternal alleles is maintained in the human, as it is in the mouse  and a small fraction of the human population may have transcriptional imprinting of IGF2R . We used a single nucleotide polymorphism within an MspI site to identify maternal and paternal alleles of informative (heterozygous) individuals and a methylation-sensitive restriction endonuclease (NotI) to determine the methylation status of a specific CpG site within the DMR . We identified 28 in vitro and 27 in vivo individuals who were informative and assayed allele-specific methylation as described previously .
We calculated the ratio between the DNA methylation levels on paternal and maternal (P/M) alleles as an indicator of methylation imprint status. A ratio of zero corresponds to exclusive methylation of the NotI site on the maternal allele, while a ratio of one signifies methylation of this site on an equal number of paternal and maternal alleles. In the case of controls, no uncleaved C alleles were detected in any C/C homozygous individuals, indicating that MspI cleaved the PCR products with >99% efficiency.
Although preferential methylation of the presumed maternal allele was observed in almost all individuals (Figure 3), the distribution of paternal/maternal (P/M) methylation ratios at the IGF2R DMR in cord blood and cord showed that most individuals have an easily measurable level of methylation at the CpG within the NotI cleavage site on the presumed paternal allele (P/M>0.1, Figure 3), as has also been observed previously in peripheral blood from the CEPH families .
(A) 27 informative in vivo individuals, and (B) 28 informative in vitro individuals. P/M represents the DNA methylation ratio between paternal and maternal alleles (see text). Each individual is represented as vertical data-set of P/M ratios from cord blood, cord and five sections of placenta. Individuals were ranked based on the degree of scatter (intra-individual variation) observed.
In cord blood, only 34% of the total population has low levels of methylation on the paternal allele (P/M allele ratios of <0.1) while a very small fraction of individuals (2%) have P/M allelic methylation ratios greater than 0.5. The distribution of individual P/M methylation ratios in cord (Figure 3) also showed a similar pattern to what was observed in cord blood. However, results from five placenta sections taken from the same individuals showed nearly complete loss of the methylation imprint (i.e. P/M ratios close to 1) at this locus in samples from multiple individuals in both in vitro and in vivo groups (Figure 3).
We found no difference in mean P/M ratios in cord blood, cord or placenta between the in vitro and in vivo groups, either comparing individual tissue types or combining all samples (Table 4). Cord blood allele-specific methylation ratio variance was greater in the in vitro group (P = 0.0016) but we did not attempt to calculate a cord blood stem cell number comparison because of the small number of samples on which to model the distribution. There was no significant difference in the population variance in the in vitro group in cord or placenta, although the presence of a substantial fraction of samples in both groups for which nearly complete loss of the methylation imprint (P/M>0.9) was observed is likely to affect our ability to distinguish such a difference.
IGF2, H19, and IGF2R Steady-State mRNA Levels
We measured steady-state IGF2, H19 and IGF2R mRNA levels in cord blood and placenta from children conceived in vitro or in vivo (Table 5, Figure 4). In addition to the children who were informative for allele specific DMR methylation (Figure 1 and Figure 3), we also measured mRNA levels in the children who were not informative. Mean cord blood IGF2R mRNA levels were significantly lower in the in vitro group (fold change = 0.61, P = 0.0039). Mean placental IGF2 and H19 mRNA levels were also significantly lower in the in vitro group (fold change = 0.52, P<0.0001, and fold change = 0.72, P = 0.0193, respectively).
Each symbol represents the mRNA level in one individual. (A) H19 in cord blood (in vitro n = 73, in vivo n = 118, fold change 0.88, P = 0.8774), (B) IGF2 in cord blood (in vitro n = 77, in vivo n = 116, fold change 1.03, P = 0.5551), (C) IGF2R in cord blood (in vitro n = 75, in vivo n = 121, fold change 0.61, P = 0.0039), (D) H19 in placenta (in vitro n = 84, in vivo n = 135, fold change 0.72, P = 0.0193), (E) IGF2 in placenta (in vitro n = 98, in vivo n = 160, fold change 0.52, P<0.0001), and (F) IGF2R in placenta (in vitro n = 97, in vivo n = 148, fold change 1.18, P = 0.2227).
We have examined intra- and inter-individual variation in DNA methylation at the differentially methylated regions (DMRs) of the IGF2/H19 and IGF2R loci in cord blood, cord and five sections of placenta from a population of children conceived in vitro or in vivo. Although a significant fraction of individuals in both groups do appear to maintain the IGF2/H19 methylation imprint “correctly”, with M/P ratios of <0.1 (93.6% of peripheral blood samples are below this M/P ratio, Sandovici et al., 2003) , we found substantial intra-individual and inter-individual variation in allele-specific methylation in both groups in all three tissues: 8/46 individuals in the in vivo group and 7/40 individuals in the in vitro group have cord blood M/P ratios above this level (n.b. of the 56 informative in vivo and 45 informative in vitro individuals shown in Figure 1, cord blood samples were unavailable for 10 of the in vivo and five of the in vitro children). The acquisition of CpG methylation on only a fraction of maternal IGF2/H19 DMRs in so many individuals suggests that there is extensive population level variation in the time at which methylation imprints become set during development, especially in extraembryonic lineages. This assertion receives further support from the analysis of intra-individual variation in P/M methylation ratios at an IGF2R DMR. Very few individuals (one in the in vivo group, two in the in vitro group) maintained this imprint “correctly” in all samples (even if the threshold for “correctly” is reduced to P/M ratios <0.2) and the discrepancy between maintaining the imprint in embryonic and extraembryonic tissue is even more pronounced (Figure 3). The observation of greater epigenetic variation in extraembryonic than in embryonic tissues is consistent with observations at a number of imprinted loci, including IGF2/H19, in the mouse . We believe this is the first time this observation has been made in the human. Parental allele-specific mean methylation ratios at the IGF2/H19 DMR were greater in the in vitro group, indicating acquisition of methylation on maternal alleles. This finding is consistent with observations made at Igf2/H19 in the mouse .
The total variance in M/P ratio at IGF2/H19 was significantly greater in cord blood and cord from the in vitro group, with a suggestive P-value (0.0620) for placenta. These findings indicate an association between ART and the magnitude of the variance in parental allele-specific methylation patterns. The mechanism by which greater variance might be created in the in vitro group is unclear but could be related to the number of trophoblast stem cells that give rise to the placenta in each group. If fewer trophoblast stem cells give rise to the placenta in the in vitro group, one expects greater intra-individual variation in somatically heritable epigenetic marks if the population of stem cells contains cells with more than one epigenetic state (i.e. IGF2/H19 DMRs that are methylated only on the paternal allele in some cells and methylated on both maternal and paternal alleles in others).
As an independent test of this prediction, we measured X chromosome inactivation ratios in five sections of placenta in females from the in vitro (50 individuals) and in vivo groups (54 individuals) in order to estimate the number of trophoblast stem cells that give rise to the placenta in each group. The closest fit to the in vitro X-inactivation distribution corresponds to nine trophoblast stem cells while the closest fit to the in vivo distribution corresponds to 11 trophoblast stem cells. We additionally estimated the number of trophoblast stem cells in each group by comparing the distribution of M/P IGF2/H19 DMR methylation ratios from the two comparison groups. The closest fit to the distribution of M/P ratios in placenta in children conceived in vitro corresponds to eight trophoblast stem cells and the closest fit in children conceived in vivo corresponds to 10 trophoblast stem cells. Overall, these estimates of between group intra-individual epigenetic variation are consistent with the prediction that overall greater variance in M/P IGF2/H19 DMR methylation ratios in the in vitro group is associated with fewer trophoblast stem cells.
Although our statistical estimates of the number of trophoblast stem cells is imprecise, it gives one some confidence that similar absolute numbers are obtained (Table 2) from estimates of epigenetic variance at two different loci, examining different individuals with each assay. Whether the calculated difference in number of trophoblast stem cells reflects designation of trophoblast stem cells at an earlier stage of development (when fewer cells are present in the embryos) or whether embryos from the in vitro group have fewer cells than in vivo embryos at comparable stages cannot be determined from these data, however, previous reports suggest that in vitro mouse embryos may contain fewer cells than in vivo embryos at the same developmental time . Our approach of using epigenetic variance to calculate the number of trophoblast stem cells is, for obvious reasons, the only opportunity for estimating this number in the human because direct comparisons of cell numbers in in vitro and in vivo embryos is not possible.
We note that while we have uncovered locus-specific differences in the level of epigenetic variation in children conceived in vitro, we cannot distinguish whether the differences are due to some aspect of the assisted reproduction process or is related to the underlying infertility. In fact, the characteristic of epigenetic variance, itself, may be under genetic control and may also be influenced by the environment . Greater variance in trait value, even without changes in trait mean, is predicted to have a substantial positive effect on fitness in a changing environment . In this regard, we note that in vitro conception is associated with at least two changes of environment (hormonal stimulation, retrieval of ova from the maternal environment to fertilization and culture in vitro, followed by return to the maternal environment). A larger-scale epigenetic screen is required in order to determine whether there is a tendency for in vitro conception to be associated with overall increased variance of epigenetic marks.
Several observations are noteworthy about the steady-state mRNA levels measured for IGF2 in placenta. First, IGF2 mRNA levels in placentas from in vitro conceived children, as a group, were approximately half of what was observed in children conceived in vivo. This observation is consistent with experiments demonstrating reduced Igf2 mRNA levels in placentas from mouse embryos subject to in vitro manipulations . Second, reduction in IGF2 mRNA levels in the human placentas does not occur in conjunction with loss of methylation at the paternal DMR, as expected if transcript levels are controlled by genomic imprinting alone. Furthermore, we did not observe that increased levels of methylation at the maternal DMR induced a coordinate level of transcription from the maternal IGF2 allele. In fact, given that IGF2 transcript levels vary by more than an order of magnitude between individuals (Figure 4) and almost by that much between samples within a single placenta (Figure S1), the mechanism by which natural selection might act in a population, on a process whose postulated design is to reduce transcription by half (from two alleles to one) is unclear. Along these same lines, we observed no correlation between birth weight and IGF2 transcript levels in either placenta or cord blood, whether or not birth weights were corrected for gestational age (Figure S2). This last observation was not completely unexpected, as several laboratories have failed to observe a correlation between IGF2 mRNA levels and birth weight –. In this regard, it is likely that epigenetic marking of genes according to parental origin plays an important role in other processes associated with reproduction and the formation of gametes, such as chromosome pairing and recombination –. The selective force for the maintenance of imprinting in these processes is both direct (successful recombination is required for successful gametogenesis) and related to reproductive success.
We also observed that mean steady-state levels of IGF2R mRNA were lower in cord blood from the in vitro group. This locus does not appear to be transcriptionally imprinted in most humans , , although the preferential methylation of only one parent's allele (the maternal) is conserved , . Although not transcriptionally imprinted, we did note an inverse correlation between methylation of the paternal allele and overall transcript level, indicating that “aberrant” methylation of the “incorrect”, paternal allele does have a small effect (accounting for ∼10% of the variance, Figure S3). Lower IGF2R mRNA level in the in vitro group is, on the face of it, in contrast to expectations. If IGF2R is a receptor that acts as a “sink” for IGF2 , , one might expect children conceived in vitro to have higher levels of IGF2R because they have a higher probability of low birth weight . In any case, we did not find any correlation between birth-weight, IGF2 levels, IGF2R levels or IGF2/IGF2R ratios (Figure S2).
Overall, our results indicate that epigenetic modifications at IGF2/H19 and IGF2R are subject to frequent changes during early development, especially in extraembryonic tissues. Although not all of the epigenetic changes appear to be manifested as significant differences in DNA methylation, conception in vitro is associated with gene expression differences for all three genes in some tissues. Whether the gene expression differences between in vitro and in vivo groups are also a manifestation of what appears to be a smaller number of trophoblast stem cells in children from the in vitro group is a subject for future investigation.
Materials and Methods
The cases/in vitro group are newborns conceived by assisted reproductive technology at a single infertility treatment center so that the clinical and laboratory procedures are uniform. The parents of the control/in vivo group had no prior history of infertility and the index pregnancy was achieved without medical assistance, such as the use of infertility medications or treatments. All the in vitro patients were stimulated with commercially available gonadotrophin preparations. The embryo culture media and the incubation parameters were all the same. The cases and controls were matched with regards to maternal age, race and gestational age (Table S1). Written, informed consent was obtained in advance from the mother of each newborn (University of Pennsylvania I.R.B. approved protocol no. 804530). A summary of the assays used, number of individuals studied and the tissues investigated is provided in Table S2.
Cord blood, cord and placenta samples were collected from each in vitro and in vivo newborn. All cord blood samples were collected within 20 min of delivery. Tissue samples were stored at 4°C after delivery, and samples were collected within five hours of delivery . The umbilical cord was wiped with normal saline and the cord vein was punctured with a 21G needle. Whole cord blood (6–10 ml) was collected in lavender topped vacutainer tubes at room temperature. The sample was shaken thoroughly to prevent clotting as the tube contains EDTA, ethylenediaminetetraacetic acid. An aliquot (3–4 ml) of cord blood was transferred to a 15 ml Falcon tube containing RNALater RNA Stabilization Reagent (Ambion, USA), following the manufacturers guidelines, to stabilize the RNA. The remaining cord blood in the lavender topped vacutainer tubes was saved for blood DNA extraction. All cord blood DNA and RNA samples were initially stored at 4°C, and nucleic acid extractions were performed within 2–4 days of collection.
Placental tissue (1.5–2.5 cm3) was excised from the fetal surface of the placenta and rinsed extensively with sterile saline solution to minimize maternal blood contamination. Each placenta was sampled from four quadrants and from directly behind the cord insertion site (this sample was used for the RT-PCR and pyrosequencing assays, as well as for the allele-specific methylation assays). A segment of umbilical cord (2 cm) was cut and treated in a similar fashion. Placental and cord tissue for RNA extraction were chopped into small pieces (0.5 cm3) and immersed in RNALater RNA Stabilization Reagent (Ambion, USA), following the manufacturers guidelines, as soon as possible after collection. All tissue DNA and RNA samples were initially stored at 4°C, and tissue digestion and nucleic acid extractions were performed within 2–4 days of collection. Approximately 4–5 mg of tissue was used for the DNA and RNA extraction procedures, and the remaining tissue was stored at −80°C.
DNA and RNA Isolation
Cord blood DNA was isolated using the ArchivePure DNA Blood Kit (Fisher Scientific Company, USA) following the manufacturers guidelines. Tissue genomic DNA was extracted using standard phenol-chloroform extraction methods. The isolated DNA was dissolved in 10 mM TrisCl, pH 8.0, quantified using a spectrophotometer and stored at −80°C until further use. Cord blood RNA was isolated using the PerfectPure RNA Blood Kit (Fisher Scientific Company, USA) following the manufacturers guidelines.
Total cellular RNA was extracted from each tissue sample using TRIzol Reagent (Invitrogen Corporation, USA), according to the manufacturers instructions. The isolated RNA was dissolved in Milli-Q water, quantified using a spectrophotometer and stored at −80°C until further use.
Parental Allele-Specific Methylation at IGF2/H19 and IGF2R DMRs
There are many DMRs on chromosome 11, but the most consistent observations indicating a role in the control of transcription of the IGF2 and H19 genes involve a CpG island located in a 5 kb region centromeric to the H19 gene, known as the IGF2/H19 DMR . CpG sites within this DMR on the paternal allele are normally methylated, while those on the maternal allele are normally unmethylated , –. This region also contains seven different binding sites for the CTCF protein  and the methylation status of the sixth binding site was found to be most consistently associated with the transcriptional status of both IGF2 and H19 .
The upstream H19 sequence used in this study is available from GenBank (accession number AF125183). Allele-specific methylation was investigated by screening the DNA samples for a C/T polymorphism recognized by CfoI at the IGF2/H19 DMR (near the sixth binding site for CTCF) . After identifying maternal and paternal alleles of heterozygous individuals, a methylation-sensitive restriction endonuclease (MluI) was used to determine the methylation status of specific CpG sites within the DMR. If all paternal alleles are methylated and all maternal alleles are unmethylated at these sites in a sample of genomic DNA, then all maternal alleles should be cleaved by MluI while all paternal alleles will remain uncleaved. Amplification of the region by PCR using primers that flank the MluI site should amplify only paternal alleles (identified by post-PCR cleavage with CfoI). Amplification of maternal alleles indicates resistance to cleavage by MluI. This may occur as a result of methylation of the CpG site within the MluI recognition sequence (the principle upon which the assay is based), mutation of the MluI site or technical artifact. The latter two possibilities may be distinguished from the first by DNA sequencing, assay reproducibility and use of additional methylation-sensitive restriction endonucleases.
The IGF2R DMR is located in the second intron of IGF2R and is normally methylated on the maternal allele. The sequence of IGF2R is available from GenBank (accession number AF069333). Allele-specific methylation at the IGF2R DMR was investigated by screening the DNA for a C/T polymorphism recognized by MspI. After identifying maternal and paternal alleles of heterozygous individuals, a methylation-sensitive restriction endonuclease (NotI) was used to determine the methylation status of specific CpG sites within the DMR.
Genomic DNA (100 ng) from informative individuals was digested overnight at 37°C with an excess of a methyl-sensitive restriction endonuclease: MluI and NotI for the IGF2/H19 and IGF2R DMRs, respectively. Control individuals who were homozygous for C alleles and homozygous for T alleles were also analyzed in each experiment.
After digestion, the enzymes were denatured and the digested DNA was amplified in a hot-stop PCR assay using the following primers: IGF2-F 5′-GAGATGGGAGGAGATACTAGG-3′ and IGF2-R 5′-GTCAGTTCAGTAAAAGGCTGG-3′ for the IGF2/H19 DMR, and IGF2R-F 5′-GGCCGAGGCCTGGCATGTTGG -3′ and IGF2R-R 5′-TGGGGAAGCGCGAGAGGCCTAGG-3′ for the IGF2R DMR. After 30 cycles at 94°C for 30 s, 50°C for 30 s (IGF2/H19) or 63°C for 30 s (IGF2R), and 72°C for 1 min, we added 3 µCi α-32P dCTP for one additional cycle and a final elongation step (72°C for 7 min). PCR products were then digested overnight at 37°C with the enzyme used for identifying the parental origin of the alleles (CfoI for IGF2/H19 and MspI for IGF2R). The samples were separated on denaturing 5% polyacrylamide gels and the intensity of the bands (alleles) were quantified using a PhosphoImage Reader FLA 5000 (FUJIFILM Medical Systems USA, Inc.).
Analyses of IGF2/H19 DMR Methylation by Pyrosequencing
We used a custom pyrosequencing assay for the IGF2/H19 DMR (NCBI36:11,2019856-2019740) which included five CpGs. Genomic DNA (500 ng per sample) was bisulfite treated using EZ Gold DNA Methylation Kit (Zymo Research, USA) following the manufacturers protocol.
Bisulfite treated DNA was used for generating PCR amplified templates for pyrosequencing. The PCR primer sequences were: forward 5′- GGGGTTATTTGGGAATAGG-3′ and biotin labeled reverse, 5′- CCAAACCATAACACTAAAACCCTC-3′. The PCR reaction (30 µl) was following: 25 ng of bisulfite DNA, 0.75 U HotStar Taq Polymerase (Qiagen, USA), 1X PCR buffer, 3 mM MgCl2, 200 µM of each dNTPs, 6 pmol forward primer and 6 pmol reverse primer. Recommended PCR cycling conditions were: 95°C for 15 min; 45 cycles (95°C for 30 s; 60°C for 30 s; 72°C for 30 s); 72°C for 5 min. The biotinylated PCR product (10 µl) was used for each sequencing assay with the following sequencing primer: 5′- GAATAGGATATTTATAGGAG-3′. Pyrosequencing was done using the PSQ96HS system according to standard procedures using Pyro Gold Reagent kits (Biotage, Sweden). Methylation was quantified using Pyro Q-CpG Software (Biotage, Sweden), which calculates the ratio of converted C's (T's) to unconverted C's at each CpG and expresses this as a percentage methylation.
Quantitative Real-Time RT-PCR
First-strand cDNA was obtained using SuperScript III Reverse Transcriptase (RT) (Invitrogen Corporation, USA). To produce cDNA from total RNA, a mixture containing 0.5–1 µg extracted total RNA, 0.5 µg oligo(dT)18 primer and 1 µl dNTP mix (10 mM each) in a final 13 µl reaction volume was heated to 65°C for 5 min, cooled down on ice for 1 min, and then added to a 7 µl reaction mixture containing 4 µl SuperScript III RT buffer (10), 1 µl DTT (0.1 M), 1 µl RNaseOUT Recombinant RNase inhibitor (40 U/µl; Invitrogen Corporation, USA) and 1 µl SuperScript III M-MLV reverse transcriptase (200 U/µl). The samples were mixed and incubated at 50°C for 60 min. Reactions were terminated at 70°C for 15 min and the RT products were stored at −20°C until further use.
Quantitative real-time RT-PCR assays were carried out using a 7700 Sequence Detector (Applied Biosystems, USA). GAPDH, which has previously been used as a housekeeping gene in placenta by several investigators –, was used as the housekeeping gene. All the placental tissue samples were from the third trimester. There was a positive correlation between GAPDH expression and the expression of another commonly used housekeeping gene HPRT, when studied in the same samples (Figure S4).
Steady-state mRNA levels of IGF2, H19, IGF2R and housekeeping gene GAPDH were measured using gene-specific primers and QuantiFast SYBR Green PCR Master Mix (Qiagen, USA). The primer sequences were following: IGF2 Forward 5′-TCTGACCTCCGTGCCTA-3′, IGF2 Reverse 5′-TTGGGATTGCAAGCGTTA-3′, H19 Forward 5′-AGAAGCGGGTCTGTTTCTTTA-3′, H19 Reverse 5′-TGGGTAGCACCATTTCTTTCA-3′, IGF2R Forward 5′-ACCTCAGCCGTGTGTCCTCT-3′, IGF2R Reverse 5′-CTCCTCTCCTTCTTGTAGAGCAA-3′, GAPDH Forward 5′-GAGTCAACGGATTTGGTCGT-3′, and GAPDH Reverse 5′-TTGATTTTGGAGGGATCTCG-3′. PCR reactions were performed by mixing 1 µl of cDNA (50 ng/µl placenta, 25 ng/µl cord blood) with 24 µl of reaction mixture (12.5 µl QuantiFast SYBR Green PCR Master Mix (2X), 2.5 µl forward primer (10 µM), 2.5 µl reverse primer (10 µM), and 6.5 µl nuclease free dH2O) and amplified under the following conditions: 95°C for 5 min, followed by 40 cycles of 95°C for 10 s and 60°C for 30 s. A melting curve analysis of the PCR products was performed to verify their specificity and identity. PCR products were also run on 2% agarose gels to confirm the size of the amplified products. Relative gene expression levels were obtained using the ΔΔCT method .
Quantitative Analysis of IGF2 and H19 Imprinting Status
To avoid genomic DNA contamination during imprinting analysis, PCR was done across an intron-exon boundary and the cDNA products were gel-purified. The primers used for assaying IGF2 imprinting were: primer 1, 5′−ATCGTTGAGGAGTGCTGTTTC−3′; primer 2, 5′−CGGGGATGCATAAAGTATGAG−3′; primer 3, 5′−CTTGGACTTTGAGTCAAATTGG−3′; and primer 4, 5′−GGTCGTGCCAATTACATTTCA−3′ . Heterozygosity of an ApaI polymorphism in exon 9 of IGF2 was ascertained by doing PCR with genomic DNA using primers 3 and 4 across the ApaI site, and the PCR product was digested with ApaI. Imprinting status was ascertained by doing RT-PCR, using primers 1 and 2 in exons 8 and 9, respectively. The cDNA PCR product, which is shorter than any possible contaminating genomic DNA product because of intron splicing, was electrophoresed and purified from a 2% agarose gel using the QIAquick Gel Extraction Kit (Qiagen, USA) following the manufacturers protocol. Hot-stop PCR was then done using primers 3 and 4, with α-32P dCTP added before the last cycle. The PCR product was digested with ApaI and then separated on denaturing 5% polyacrylamide gels and the intensity of the bands (alleles) were quantified using a PhosphoImage Reader FLA 5000 (FUJIFILM Medical Systems USA, Inc.).
The primers used for assaying H19 imprinting were: H19-1, 5′-GGAGTTGTGGAGACGGCCTTGAGT-3′; H19-2, 5′-CCAGTCACCCGGCCCAGATGGAG-3′; and H19-3, 5′-CTTTACAACCACTGCACTACCTGAC-3′. Heterozygosity of an RsaI polymorphism in exon 5 of H19 was ascertained by doing PCR with genomic DNA using primers H19-1 and H19-2 across the RsaI site, and the PCR product was digested with RsaI. Imprinting status was ascertained by doing RT-PCR, using primers H19-3 and H19-2 in exons 4 and 5, respectively. The cDNA PCR product, which is shorter than any possible contaminating genomic DNA product because of intron splicing, was electrophoresed and purified from a 2% agarose gel using the QIAquick Gel Extraction Kit (Qiagen, USA) following the manufacturers protocol. Hot-stop PCR was then done using primers H19-1 and H19-2, with α-32P dCTP added before the last cycle. The PCR product was digested with RsaI and then separated on denaturing 5% polyacrylamide gels and the intensity of the bands (alleles) were quantified using a PhosphoImage Reader FLA 5000 (FUJIFILM Medical Systems USA, Inc.).
X-chromosome inactivation ratios were assayed using previously published modifications of a methylation-sensitive PCR assay , , –. We measured the methylation status of a CpG site that is correlated with the expression of alleles at the X-linked, highly polymorphic androgen receptor (AR) locus. Genomic DNA from cord blood, cord and five sections of placenta was available for 50 in vitro and 54 in vivo females who were heterozygous for AR alleles that differed by more than one CAG repeat.
Genomic DNA from females who were heterozygous (informative) at the highly polymorphic (CAG)n repeat of the X-linked AR gene was amplified after previous overnight digestion with HhaI methyl-sensitive restriction endonuclease, with primers (AR1: 5′-AGAGGCCGCGAGCGCAGCAC-3′ and AR2: 5′-ACTCCAGGGCCGACTGCGGC-3′), which flank the repeat and two HhaI sites. We added a radiolabeled nucleotide for the last cycle of ‘hot-stop’ PCR, rather than a single end-labeled primer, to increase the signal. After 27 cycles at 94°C for 1 min, 68°C for 1 min, and 72°C for 1 min, 3 µCi α-32P dCTP was added for one additional cycle. PCR products were separated on denaturing 5% polyacrylamide gels and the intensity of the alleles was quantified by using the PhosphoImage Reader FLA 5000 (FUJIFILM Medical Systems USA, Inc.) . As a way of quantifying the degree of skewing, i.e., the degree to which the somatic cells of an individual female deviated from a 1∶1 ratio, the intensity of the upper allele divided by the sum of the intensities of both alleles was computed for each individual.
The statistical significance of the methylation datasets representing the in vivo and in vitro group was examined using the Wilcoxon Rank Sums Test. Data from the real time RT-PCR experiments were analyzed using Student's T-test. The number of cells was estimated using the method described by Amos-Landgraf et al. (2006) and Mclaren A (1972) , . P-values ≤0.05 were considered significant.
Intra-individual variation in IGF2 mRNA expression. Steady-state IGF2 mRNA levels measured in five sections of placenta in 12 individuals. IGF2 transcript levels vary by more than an order of magnitude (n.b.: each unit on the vertical axis is a power of 2) between individuals and between samples within some individuals.
(0.05 MB TIF)
Lack of correlation between birth weight percentiles and IGF2 or IGF2R mRNA levels in cord blood and placenta. Birth weight percentiles in the in vitro and in vivo populations plotted versus cord blood transcript levels of (A) IGF2, (B) IGF2R, (C) IGF2/IGF2R, and placenta transcript levels of (D) IGF2, (E) IGF2R, and (F) IGF2/IGF2R. Although average birth weight was lower in the in vitro group, we found little correlation between birth weight and IGF2 or IGF2R transcript levels (regression lines are shown, straight and dashed lines for in vivo and in vitro populations, respectively, and the maximum r2of 0.2433 is found in (B), even when birth weights were corrected for gestational age [Min, et al; Oken, et al; Yarkoni, et al]. These findings are consistent with previous observations from studies in human IUGR placentae . [Min SJ, Luke B, Min L, Misiunas R, Nugent C, et al. (2004) Birth weight references for triplets. Am J Obstet Gynecol 191(3): 809-814. Oken E, Kleinman KP, Rich-Edwards J, Gillman MW (2003) A nearly continuous measure of birth weight for gestational age using a United States national reference. BMC Pediatr 8(3): 6. Yarkoni S, Reece EA, Holford T, O'Connor TZ, Hobbins JC (1987) Estimated fetal weight in the evaluation of growth in twin gestations: a prospective longitudinal study. Obstet Gynecol 69(4): 636-639.)
(0.46 MB TIF)
Correlation between IGF2R expression and P/M methylation ratios in cord blood.
(0.06 MB TIF)
Positive correlation between GAPDH expression and the expression of another commonly used housekeeping gene, HPRT, when studied in the same placenta samples.
(0.12 MB TIF)
(0.04 MB DOC)
Summary of assays used, number of individuals studied, and tissues investigated.
(0.05 MB DOC)
The authors thank Halleh Asadpour, Noa Holtzman, Poornima Ramkumar, Lisa Korostowski, Siddhartha Kosaraju, and Daryoush Saeed-Vafa for technical assistance and sample preparation.
Conceived and designed the experiments: NT SK RC CC CS. Performed the experiments: NT SK LFG. Analyzed the data: NT SK JPG CC CS. Contributed reagents/materials/analysis tools: NT SK RC MWF JPG. Wrote the paper: NT SK CC CS.
- 1. Ludwig M, Katalinic A, Gross S, Sutcliffe A, Varon R, et al. (2005) Increased prevalence of imprinting defects in patients with Angelman syndrome born to subfertile couples. J Med Genet 42: 289–291.M. LudwigA. KatalinicS. GrossA. SutcliffeR. Varon2005Increased prevalence of imprinting defects in patients with Angelman syndrome born to subfertile couples.J Med Genet42289291
- 2. Orstavik KH, Eiklid K, van der Hagen CB, Spetalen S, Kierulf K, et al. (2003) Another case of imprinting defect in a girl with Angelman syndrome who was conceived by intracytoplasmic semen injection. Am J Hum Genet 72: 218–219.KH OrstavikK. EiklidCB van der HagenS. SpetalenK. Kierulf2003Another case of imprinting defect in a girl with Angelman syndrome who was conceived by intracytoplasmic semen injection.Am J Hum Genet72218219
- 3. Cox GF, Burger J, Lip V, Mau UA, Sperling K, et al. (2002) Intracytoplasmic sperm injection may increase the risk of imprinting defects. Am J Hum Genet 71: 162–164.GF CoxJ. BurgerV. LipUA MauK. Sperling2002Intracytoplasmic sperm injection may increase the risk of imprinting defects.Am J Hum Genet71162164
- 4. Manipalviratn S, DeCherney A, Segars J (2009) Imprinting disorders and assisted reproductive technology. Fertil Steril 91: 305–315.S. ManipalviratnA. DeCherneyJ. Segars2009Imprinting disorders and assisted reproductive technology.Fertil Steril91305315
- 5. Lim D, Bowdin SC, Tee L, Kirby GA, Blair E, et al. (2009) Clinical and molecular genetic features of Beckwith-Wiedemann syndrome associated with assisted reproductive technologies. Hum Reprod 24: 741–747.D. LimSC BowdinL. TeeGA KirbyE. Blair2009Clinical and molecular genetic features of Beckwith-Wiedemann syndrome associated with assisted reproductive technologies.Hum Reprod24741747
- 6. Bowdin S, Allen C, Kirby G, Brueton L, Afnan M, et al. (2007) A survey of assisted reproductive technology births and imprinting disorders. Hum Reprod 22: 3237–3240.S. BowdinC. AllenG. KirbyL. BruetonM. Afnan2007A survey of assisted reproductive technology births and imprinting disorders.Hum Reprod2232373240
- 7. Hansen M, Bower C, Milne E, de Klerk N, Kurinczuk JJ, et al. (2005) Assisted reproductive technologies and the risk of birth defects - a systematic review. Hum Reprod 20: 328–338.M. HansenC. BowerE. MilneN. de KlerkJJ Kurinczuk2005Assisted reproductive technologies and the risk of birth defects - a systematic review.Hum Reprod20328338
- 8. Halliday J, Oke K, Breheny S, Algar E, J Amor D (2004) Beckwith-Weidemann Syndrome and IVF: a case-control study. Am J Hum Genet 75: 526–528.J. HallidayK. OkeS. BrehenyE. AlgarD. J Amor2004Beckwith-Weidemann Syndrome and IVF: a case-control study.Am J Hum Genet75526528
- 9. Stouder C, Deutsch S, Paoloni-Giacobino A (2009) Superovulation in mice alters the methylation pattern of imprinted genes in the sperm of the offspring. Reprod Toxicol 28: 536–541.C. StouderS. DeutschA. Paoloni-Giacobino2009Superovulation in mice alters the methylation pattern of imprinted genes in the sperm of the offspring.Reprod Toxicol28536541
- 10. Sato A, Otsu E, Negishi H, Utsunomiya T, Arima T (2007) Aberrant DNA methylation of imprinted loci in superovulated oocytes. Hum Reprod 22: 26–35.A. SatoE. OtsuH. NegishiT. UtsunomiyaT. Arima2007Aberrant DNA methylation of imprinted loci in superovulated oocytes.Hum Reprod222635
- 11. Mann MR, Lee SS, Doherty AS, Verona RI, Nolen LD, et al. (2004) Selective loss of imprinting in the placenta following preimplantation development in culture. Development 131: 3727–3735.MR MannSS LeeAS DohertyRI VeronaLD Nolen2004Selective loss of imprinting in the placenta following preimplantation development in culture.Development13137273735
- 12. Young LE, Fernandes K, McEvoy TG, Butterwith SC, Gutierrez CG, et al. (2001) Epigenetic change in IGF2R is associated with fetal overgrowth after sheep embryo culture. Nat Genet 27: 153–154.LE YoungK. FernandesTG McEvoySC ButterwithCG Gutierrez2001Epigenetic change in IGF2R is associated with fetal overgrowth after sheep embryo culture.Nat Genet27153154
- 13. Doherty AS, Mann MR, Tremblay KD, Bartolomei MS, Schultz RM (2000) Differential effects of culture on imprinted H19 expression in the preimplantation mouse embryo. Biol Reprod 62: 1526–1535.AS DohertyMR MannKD TremblayMS BartolomeiRM Schultz2000Differential effects of culture on imprinted H19 expression in the preimplantation mouse embryo.Biol Reprod6215261535
- 14. Lucifero D, Mann MR, Bartolomei MS, Trasler JM (2004) Gene-specific timing and epigenetic memory in oocyte imprinting. Hum Mol Genet 13: 839–849.D. LuciferoMR MannMS BartolomeiJM Trasler2004Gene-specific timing and epigenetic memory in oocyte imprinting.Hum Mol Genet13839849
- 15. Kerjean A, Dupont JM, Vasseur C, Le Tessier D, Cuisset L, et al. (2000) Establishment of the paternal methylation imprint of the human H19 and MEST/PEG1 genes during spermatogenesis. Hum Mol Genet 9: 2183–2187.A. KerjeanJM DupontC. VasseurD. Le TessierL. Cuisset2000Establishment of the paternal methylation imprint of the human H19 and MEST/PEG1 genes during spermatogenesis.Hum Mol Genet921832187
- 16. Swales AK, Spears N (2005) Genomic imprinting and reproduction. Reproduction 130(4): 389–399.AK SwalesN. Spears2005Genomic imprinting and reproduction.Reproduction130(4)389399
- 17. Reik W, Dean W, Walter J (2001) Epigenetic reprogramming in mammalian development. Science 293: 1089–1093.W. ReikW. DeanJ. Walter2001Epigenetic reprogramming in mammalian development.Science29310891093
- 18. Fortier AL, Lopes FL, Darricarrère N, Martel J, Trasler JM (2008) Superovulation alters the expression of imprinted genes in the midgestation mouse placenta. Hum Mol Genet 17: 1653–1665.AL FortierFL LopesN. DarricarrèreJ. MartelJM Trasler2008Superovulation alters the expression of imprinted genes in the midgestation mouse placenta.Hum Mol Genet1716531665
- 19. Market-Velker BA, Zhang L, Magri LS, Bonvissuto AC, Mann MR (2010) Dual effects of superovulation: loss of maternal and paternal imprinted methylation in a dose-dependent manner. Hum Mol Genet 19(1): 36–51.BA Market-VelkerL. ZhangLS MagriAC BonvissutoMR Mann2010Dual effects of superovulation: loss of maternal and paternal imprinted methylation in a dose-dependent manner.Hum Mol Genet19(1)3651
- 20. Rivera RM, Stein P, Weaver JR, Mager J, Schultz RM, et al. (2008) Manipulations of mouse embryos prior to implantation result in aberrant expression of imprinted genes on day 9.5 of development. Hum Mol Genet 17: 1–14.RM RiveraP. SteinJR WeaverJ. MagerRM Schultz2008Manipulations of mouse embryos prior to implantation result in aberrant expression of imprinted genes on day 9.5 of development.Hum Mol Genet17114
- 21. Fernandez-Gonzalez R, Moreira P, Bilbao A, Jimenez A, Perez-Crespo M, et al. (2004) Long-term effect of in vitro culture of mouse embryos with serum on mRNA expression of imprinting genes, development, and behavior. Proc Natl Acad Sci USA 101: 5880–5885.R. Fernandez-GonzalezP. MoreiraA. BilbaoA. JimenezM. Perez-Crespo2004Long-term effect of in vitro culture of mouse embryos with serum on mRNA expression of imprinting genes, development, and behavior.Proc Natl Acad Sci USA10158805885
- 22. Khosla S, Dean W, Brown D, Reik W, Feil R (2001) Culture of preimplantation mouse embryos affects fetal development and the expression of imprinted genes. Biol Reprod 64: 918–926.S. KhoslaW. DeanD. BrownW. ReikR. Feil2001Culture of preimplantation mouse embryos affects fetal development and the expression of imprinted genes.Biol Reprod64918926
- 23. Li T, Vu TH, Ulaner GA, Littman E, Ling JQ, et al. (2005) IVF results in de novo DNA methylation and histone methylation at an Igf2-H19 imprinting epigenetic switch. Mol Hum Reprod 11: 631–640.T. LiTH VuGA UlanerE. LittmanJQ Ling2005IVF results in de novo DNA methylation and histone methylation at an Igf2-H19 imprinting epigenetic switch.Mol Hum Reprod11631640
- 24. Takai D, Gonzales FA, Tsai YC, Thayer MJ, Jones PA (2001) Large scale mapping of methylcytosines in CTCF-binding sites in the human H19 promoter and aberrant hypomethylation in human bladder cancer. Hum Mol Genet 10: 2619–2626.D. TakaiFA GonzalesYC TsaiMJ ThayerPA Jones2001Large scale mapping of methylcytosines in CTCF-binding sites in the human H19 promoter and aberrant hypomethylation in human bladder cancer.Hum Mol Genet1026192626
- 25. Srivastava M, Hsieh S, Grinberg A, Williams-Simons L, Huang SP, et al. (2000) H19 and Igf2 monoallelic expression is regulated in two distinct ways by a shared cis acting regulatory region upstream of H19. Genes Dev 14: 1186–1195.M. SrivastavaS. HsiehA. GrinbergL. Williams-SimonsSP Huang2000H19 and Igf2 monoallelic expression is regulated in two distinct ways by a shared cis acting regulatory region upstream of H19.Genes Dev1411861195
- 26. Smrzka OW, Fae I, Stoger R, Kurzbauer R, Fischer GF, et al. (1995) Conservation of a maternal-specific methylation signal at the human IGF2R locus. Hum Mol Genet 4: 1945–1952.OW SmrzkaI. FaeR. StogerR. KurzbauerGF Fischer1995Conservation of a maternal-specific methylation signal at the human IGF2R locus.Hum Mol Genet419451952
- 27. Xu Y, Goodyer CG, Deal C, Polychronakos C (1993) Functional polymorphism in the parental imprinting of the human IGF2R gene. Biochem Biophys Res Commun 197: 747–754.Y. XuCG GoodyerC. DealC. Polychronakos1993Functional polymorphism in the parental imprinting of the human IGF2R gene.Biochem Biophys Res Commun197747754
- 28. Sandovici I, Leppert M, Hawk PR, Suarez A, Linares Y, et al. (2003) Familial aggregation of abnormal methylation of parental alleles at the IGF2/H19 and IGF2R differentially methylated regions. Hum Mol Genet 12: 1569–1578.I. SandoviciM. LeppertPR HawkA. SuarezY. Linares2003Familial aggregation of abnormal methylation of parental alleles at the IGF2/H19 and IGF2R differentially methylated regions.Hum Mol Genet1215691578
- 29. Constância M, Pickard B, Kelsey G, Reik W (1998) Imprinting mechanisms. Genome Res 8(9): 881–900.M. ConstânciaB. PickardG. KelseyW. Reik1998Imprinting mechanisms.Genome Res8(9)881900
- 30. Giannoukakis N, Deal C, Paquette J, Goodyer CG, Polychronakos C (1993) Parental genomic imprinting of the human IGF2 gene. Nat Genet 4: 98–101.N. GiannoukakisC. DealJ. PaquetteCG GoodyerC. Polychronakos1993Parental genomic imprinting of the human IGF2 gene.Nat Genet498101
- 31. Ulaner GA, Vu TH, Li T, Hu JF, Yao XM, et al. (2003) Loss of imprinting of IGF2 and H19 in osteosarcoma is accompanied by reciprocal methylation changes of a CTCF-binding site. Hum Mol Genet 12: 535–549.GA UlanerTH VuT. LiJF HuXM Yao2003Loss of imprinting of IGF2 and H19 in osteosarcoma is accompanied by reciprocal methylation changes of a CTCF-binding site.Hum Mol Genet12535549
- 32. Cui H, Horon IL, Ohlsson R, Hamilton SR, Feinberg AP (1998) Loss of imprinting in normal tissue of colorectal cancer patients with microsatellite instability. Nat Med 4: 1276–1280.H. CuiIL HoronR. OhlssonSR HamiltonAP Feinberg1998Loss of imprinting in normal tissue of colorectal cancer patients with microsatellite instability.Nat Med412761280
- 33. Uejima H, Lee MP, Cui H, Feinberg AP (2000) Hot-stop PCR: a simple and general assay for linear quantification of allele ratios. Nat Genet 25: 375–376.H. UejimaMP LeeH. CuiAP Feinberg2000Hot-stop PCR: a simple and general assay for linear quantification of allele ratios.Nat Genet25375376
- 34. Amos-Landgraf JM, Cottle A, Plenge RM, Friez M, Schwartz CE, et al. (2006) X chromosome-inactivation patterns of 1,005 phenotypically unaffected females. Am J Hum Genet 79(3): 493–499.JM Amos-LandgrafA. CottleRM PlengeM. FriezCE Schwartz2006X chromosome-inactivation patterns of 1,005 phenotypically unaffected females.Am J Hum Genet79(3)493499
- 35. Mclaren A (1972) Biological Sciences: Numerology of Development. Nature 239: 274–276.A. Mclaren1972Biological Sciences: Numerology of Development.Nature239274276
- 36. Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM, Belmont JW (1992) Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgen-receptor gene correlates with X chromosome inactivation. Am J Hum Genet 51: 1229–1239.RC AllenHY ZoghbiAB MoseleyHM RosenblattJW Belmont1992Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgen-receptor gene correlates with X chromosome inactivation.Am J Hum Genet5112291239
- 37. Schoenherr CJ, Levorse JM, Tilghman SM (2003) CTCF maintains differential methylation at the Igf2/H19 locus. Nat Genet 33: 66–69.CJ SchoenherrJM LevorseSM Tilghman2003CTCF maintains differential methylation at the Igf2/H19 locus.Nat Genet336669
- 38. Hark AT, Schoenherr CJ, Katz DJ, Ingram RS, Levorse JM, et al. (2000) CTCF mediates methylation-sensitive enhancer-blocking activity at the H19/Igf2 locus. Nature 405: 486–489.AT HarkCJ SchoenherrDJ KatzRS IngramJM Levorse2000CTCF mediates methylation-sensitive enhancer-blocking activity at the H19/Igf2 locus.Nature405486489
- 39. Joyce JA, Lam WK, Catchpoole DJ, Jenks P, Reik W, et al. (1997) Imprinting of IGF2 and H19: lack of reciprocity in sporadic Beckwith-Wiedemann syndrome. Hum Mol Genet 6(9): 1543–1548.JA JoyceWK LamDJ CatchpooleP. JenksW. Reik1997Imprinting of IGF2 and H19: lack of reciprocity in sporadic Beckwith-Wiedemann syndrome.Hum Mol Genet6(9)15431548
- 40. Kalscheuer VM, Mariman EC, Schepens MT, Rehder H, Ropers HH (1993) The insulin-like growth factor type-2 receptor gene is imprinted in the mouse but not in humans. Nat Genet 5: 74–78.VM KalscheuerEC MarimanMT SchepensH. RehderHH Ropers1993The insulin-like growth factor type-2 receptor gene is imprinted in the mouse but not in humans.Nat Genet57478
- 41. Watkins AJ, Platt D, Papenbrock T, Wilkins A, Eckert JJ, et al. (2007) Mouse embryo culture induces changes in postnatal phenotype including raised systolic blood pressure. Proc Natl Acad Sci USA 104: 5449–5454.AJ WatkinsD. PlattT. PapenbrockA. WilkinsJJ Eckert2007Mouse embryo culture induces changes in postnatal phenotype including raised systolic blood pressure.Proc Natl Acad Sci USA10454495454
- 42. Feinberg AP, Irizarry RA (2009) Evolution in health and medicine Sackler colloquium: Stochastic epigenetic variation as a driving force of development, evolutionary adaptation, and disease. Proc Natl Acad Sci U S A 107(1): 1757–1764.AP FeinbergRA Irizarry2009Evolution in health and medicine Sackler colloquium: Stochastic epigenetic variation as a driving force of development, evolutionary adaptation, and disease.Proc Natl Acad Sci U S A107(1)17571764
- 43. Antonazzo P, Alvino G, Cozzi V, Grati FR, Tabano S, et al. (2008) Placenta IGF2 Expression in Normal and Intrauterine Growth Restricted (IUGR) Pregnancies. Placenta 29: 99–101.P. AntonazzoG. AlvinoV. CozziFR GratiS. Tabano2008Placenta IGF2 Expression in Normal and Intrauterine Growth Restricted (IUGR) Pregnancies.Placenta2999101
- 44. Katagiri Y, Aoki C, Fukuda Y, Takeshita N, Tanaka M, et al. (2008) Imprinted gene expression analysis in cases with low birth weight after assisted reproduction technology. Fertil Steril 90: S71.Y. KatagiriC. AokiY. FukudaN. TakeshitaM. Tanaka2008Imprinted gene expression analysis in cases with low birth weight after assisted reproduction technology.Fertil Steril90S71
- 45. Apostolidou S, Abu-Amero S, O'Donoghue K, Frost J, Olafsdottir O, et al. (2007) Elevated placental expression of the imprinted PHLDA2 gene is associated with low birth weight. J Mol Med 85: 379–387.S. ApostolidouS. Abu-AmeroK. O'DonoghueJ. FrostO. Olafsdottir2007Elevated placental expression of the imprinted PHLDA2 gene is associated with low birth weight.J Mol Med85379387
- 46. Pardo Manuel de Villena F, de la Casa-Esperón , Sapienza C (2000) Natural selection and the function of genome imprinting: beyond the silenced minority. Trends Genet 16: 573–579.F. Pardo Manuel de Villenade la Casa-EsperónC. Sapienza2000Natural selection and the function of genome imprinting: beyond the silenced minority.Trends Genet16573579
- 47. Lercher MJ, Hurst LD (2003) Imprinted chromosomal regions of the human genome have unusually high recombination rates. Genetics 165: 1629–1632.MJ LercherLD Hurst2003Imprinted chromosomal regions of the human genome have unusually high recombination rates.Genetics16516291632
- 48. Sandovici I, Kassovska-Bratinova S, Vaughan JE, Stewart R, Leppert M, et al. (2006) Human imprinted chromosomal regions are historical hot-spots of recombination. PLoS Genet 2: e101.I. SandoviciS. Kassovska-BratinovaJE VaughanR. StewartM. Leppert2006Human imprinted chromosomal regions are historical hot-spots of recombination.PLoS Genet2e101
- 49. Paigen K, Petkov P (2010) Mammalian recombination hot spots: properties, control and evolution. Nat Rev Genet 11: 221–233.K. PaigenP. Petkov2010Mammalian recombination hot spots: properties, control and evolution.Nat Rev Genet11221233
- 50. Monk D, Arnaud P, Apostolidou S, Hills FA, Kelsey G, et al. (2006) Limited evolutionary conservation of imprinting in the human placenta. PNAS 103(17): 6623–6628.D. MonkP. ArnaudS. ApostolidouFA HillsG. Kelsey2006Limited evolutionary conservation of imprinting in the human placenta.PNAS103(17)66236628
- 51. Brown J, Jones EY, Forbes BE (2009) Interactions of IGF-II with the IGF2R/cation-independent mannose-6-phosphate receptor mechanism and biological outcomes. Vitam Horm 80: 699–719.J. BrownEY JonesBE Forbes2009Interactions of IGF-II with the IGF2R/cation-independent mannose-6-phosphate receptor mechanism and biological outcomes.Vitam Horm80699719
- 52. Jones JI, Clemmons DR (1995) Insulin-like growth factors and their binding proteins:biological actions. Endocr Rev 16: 3–34.JI JonesDR Clemmons1995Insulin-like growth factors and their binding proteins:biological actions.Endocr Rev16334
- 53. Ong K, Kratzsch J, Kiess W, Costello M, Scott C, et al. (2000) Size at birth and cord blood levels of insulin, insulin-like growth factor I (IGF-I), IGF-II, IGF-binding protein-1 (IGFBP-1), IGFBP-3, and the soluble IGF-II/mannose-6-phosphate receptor in term human infants. The ALSPAC Study Team. Avon Longitudinal Study of Pregnancy and Childhood. J Clin Endocrinol Metab 85: 4266–4269.K. OngJ. KratzschW. KiessM. CostelloC. Scott2000Size at birth and cord blood levels of insulin, insulin-like growth factor I (IGF-I), IGF-II, IGF-binding protein-1 (IGFBP-1), IGFBP-3, and the soluble IGF-II/mannose-6-phosphate receptor in term human infants. The ALSPAC Study Team. Avon Longitudinal Study of Pregnancy and Childhood.J Clin Endocrinol Metab8542664269
- 54. Fajardy I, Moitrot E, Vambergue A, Vandersippe-Millot M, Deruelle P, et al. (2009) Time course analysis of RNA stability in human placenta. BMC Mol Biol 10: 21.I. FajardyE. MoitrotA. VambergueM. Vandersippe-MillotP. Deruelle2009Time course analysis of RNA stability in human placenta.BMC Mol Biol1021
- 55. Ohlsson R, Nyström A, Pfeifer-Ohlsson S, Töhönen V, Hedborg F, et al. (1993) IGF2 is parentally imprinted during human embryogenesis and in the Beckwith-Wiedemann syndrome. Nat Genet 4(1): 94–97.R. OhlssonA. NyströmS. Pfeifer-OhlssonV. TöhönenF. Hedborg1993IGF2 is parentally imprinted during human embryogenesis and in the Beckwith-Wiedemann syndrome.Nat Genet4(1)9497
- 56. Reik W, Allen ND (1994) Genomic imprinting. Imprinting with and without methylation. Curr Biol 4(2): 145–147.W. ReikND Allen1994Genomic imprinting. Imprinting with and without methylation.Curr Biol4(2)145147
- 57. Steenman MJ, Rainier S, Dobry CJ, Grundy P, Horon IL, et al. (1994) Loss of imprinting of IGF2 is linked to reduced expression and abnormal methylation of H19 in Wilms' tumour. Nat Genet 7(3): 433–439.MJ SteenmanS. RainierCJ DobryP. GrundyIL Horon1994Loss of imprinting of IGF2 is linked to reduced expression and abnormal methylation of H19 in Wilms' tumour.Nat Genet7(3)433439
- 58. Zhang Y, Tycko B (1992) Monoallelic expression of the human H19 gene. Nat Genet 1: 40–44.Y. ZhangB. Tycko1992Monoallelic expression of the human H19 gene.Nat Genet14044
- 59. Katagiri Y, Aoki C, Fukuda Y, Takeshita N, Tanaka M, et al. (2008) Imprinted gene expression analysis in cases with low birth weight after assisted reproduction technology. Fertil Steril 90(1): S7.Y. KatagiriC. AokiY. FukudaN. TakeshitaM. Tanaka2008Imprinted gene expression analysis in cases with low birth weight after assisted reproduction technology.Fertil Steril90(1)S7
- 60. Murthi P, Fitzpatrick E, Borg AJ, Donath S, Brennecke SP, et al. (2008) GAPDH, 18S rRNA and YWHAZ are suitable endogenous reference genes for relative gene expression studies in placental tissues from human idiopathic fetal growth restriction. Placenta 29(9): 798–801.P. MurthiE. FitzpatrickAJ BorgS. DonathSP Brennecke2008GAPDH, 18S rRNA and YWHAZ are suitable endogenous reference genes for relative gene expression studies in placental tissues from human idiopathic fetal growth restriction.Placenta29(9)798801
- 61. Murthi P, Doherty V, Said J, Donath S, Brennecke SP, et al. (2006) Homeobox gene HLX1 expression is decreased in idiopathic human fetal growth restriction. Am J Pathol 168(2): 511–518.P. MurthiV. DohertyJ. SaidS. DonathSP Brennecke2006Homeobox gene HLX1 expression is decreased in idiopathic human fetal growth restriction.Am J Pathol168(2)511518
- 62. Winer J, Jung CK, Shackel I, Williams PM (1999) Development and validation of real-time quantitative reverse transcriptase-polymerase chain reaction for monitoring gene expression in cardiac myocytes in vitro. Anal Biochem 270: 41–44.J. WinerCK JungI. ShackelPM Williams1999Development and validation of real-time quantitative reverse transcriptase-polymerase chain reaction for monitoring gene expression in cardiac myocytes in vitro.Anal Biochem2704144
- 63. Sandovici I, Naumova AK, Leppert M, Linares Y, Sapienza C (2004) A longitudinal study of X-inactivation ratio in human females. Hum Genet 115(5): 387–392.I. SandoviciAK NaumovaM. LeppertY. LinaresC. Sapienza2004A longitudinal study of X-inactivation ratio in human females.Hum Genet115(5)387392
- 64. Naumova AK, Olien L, Bird LM, Smith M, Verner AE, et al. (1998) Genetic mapping of X-linked loci involved in skewing of X chromosome inactivation in the human. Eur J Hum Genet 6(6): 552–562.AK NaumovaL. OlienLM BirdM. SmithAE Verner1998Genetic mapping of X-linked loci involved in skewing of X chromosome inactivation in the human.Eur J Hum Genet6(6)552562
- 65. Naumova AK, Plenge RM, Bird LM, Leppert M, Morgan K, Willard HF, Sapienza C (1996) Heritability of X chromosome-inactivation phenotype in a large family. Am J Hum Genet 58(6): 1111–1119.AK NaumovaRM PlengeLM BirdM. LeppertK. MorganHF WillardC. Sapienza1996Heritability of X chromosome-inactivation phenotype in a large family.Am J Hum Genet58(6)11111119