Abstract
In vitro fertilisation (IVF) potentially provides a profoundly abnormal environment for an embryo. Studies with mice, sheep and cattle have indicated that the culture environment of the embryo can affect the imprinting of genes and the phenotype of the animal. Recent studies have suggested that IVF causes a small but increased risk of epigenetic imprinting aberrations such as Angelman syndrome and Beckwith-Wiedemann syndrome. Our previously published IVF cohort is taller, has higher levels of growth hormones and a better lipid profile than age-matched controls. Mosaicism for the imprinting defect at SNRPN has been observed in Angelman syndrome. We hypothesised that mosaic imprinting defects may be present in phenotypically normal individuals conceived using IVF.
DNA samples from peripheral blood were obtained from 66 IVF-conceived children and 69 matched controls. DNA methylation of CpG sites within the Beckwith-Wiedemann syndrome region (H19, KCNQ1OT1 and IGF2) and the Angelman syndrome region (SNRPN) were quantified using methylation-sensitive restriction digest followed by real-time quantitative PCR (MSQ-PCR). Global DNA methylation was also examined by using MSQ-PCR on Satellite 2 repeats. No differences in the percentage of methylation between the IVF and naturally conceived children were observed at H19 (P = 0.75; unpaired t-test), KCNQ1OT1 (P = 0.98), SNRPN (P = 0.33), IGF2 (P = 0.44) or Satellite 2 (P = 0.79). These results were confirmed using bisulfite sequencing.
An individual with Prader-Willi syndrome was identified during the recruitment of this cohort. Five Prader-Willi syndrome cases have previously been identified in the IVF population. The underlying cause of Prader-Willi syndrome was identified to be a deletion of the chromosome 15q11-q13 region. This case did not provide evidence that aberrant methylation can occur during IVF.
Pyrosequencing technology was used to measure the methylation at multiple CpG sites within H19. No methylation defects were identified at H19 in the IVF group compared to naturally conceived controls. This technology proved to be prone to inaccuracies and was not used for subsequent analyses.
Genome-wide methylation analysis was examined using microarray technology and methylated DNA immunoprecipitation (MeDIP). Thirteen candidate differentially methylated genes between the IVF-conceived and control children were identified. Detailed examination of candidate genes using the Sequenom MassARRAY® EpiTYPER® system did not reveal any differential methylation at these genes assessed in the IVF and naturally conceived children. Although anthropomorphic and endocrinological data suggested a phenotypic difference between IVF and naturally conceived children, no differentially methylated genes were identified that could account for these differences.
We concluded that low-level imprinting errors are not a common occurrence in children conceived using IVF. Our data also provides reassurance that IVF-associated epigenetic errors are sporadic and rare.