Abstract
Diagnostic testing for genetic disorders or techniques such as preimplantation genetic diagnosis (PGD) both require accurate PCR genotyping (1). The failure of amplification of one allele, referred to as allelic drop-out (ADO) can confound genotyping results by falsely identifying heterozygotes as homozygous (2). A unique ADO mechanism has previously been demonstrated to occur consistently in the imprinted MEST gene, where both DNA methylation and G-quadruplex (G4) DNA structure contributed to allele loss (3). G4s are alternative DNA structures that form in G-rich regions due to the self-associating ability of guanine. Under certain ionic conditions, four guanine residues bind together either within or between strands to form a G-quartet, which can then stack upon one another to form the higher order structure. Such structures have the ability to act as a steric block to Taq polymerase. This effect is exacerbated when the G4 is methylated due to an increased thermal stability (4).
This thesis explored the hypothesis that ADO via this mechanism occurs more widely throughout the imprinted genome. To test this, 22 target loci containing G4-DNA motifs were selected from 16 imprinted genes and an assay designed to detect ADO during PCR was developed. This required the creation of two variant alleles via the introduction of a single nucleotide polymorphism (SNP) with differential primer design. Both variants were then subjected to in vitro methylation and template mixing PCR experiments followed by Sanger sequencing to reveal mono-allelic or bi-allelic amplification. Of the 22 amplicons initially selected, only 14 were able to be consistently amplified and were thus used for this analysis. This method revealed that MEST is not alone in being susceptible to ADO events, with nine other amplicons showing either complete or partial mono-allelic amplification when methylated G4s were present. To confirm that the predicted G4 motifs did adopt the structure, CD spectroscopy was used. This revealed that these motifs were capable of forming the secondary structure and therefore contributing to ADO events. This work confirms that the effect of cytosine methylation and G4 regions on ADO that was previously observed (3) occurs more widely throughout the imprinted genome, and further highlights the need for diligence in both a diagnostic and research setting when analysing imprinted genes or other methylated regions.