Quantitative one-step DNA methylation analysis using native genomic DNA as template

Abstract

During the last few years, analysis of DNA methylation has become a standard application in both research and diagnostics. Analysis of DNA methylation currently requires multi-step procedures that are either based on conversion of unmethylated cytosines by bisulfite or on methylation-sensitive endonucleases. In the present study, we investigated the potential of a novel one-step approach we refer to as quantitative one-step DNA methylation analysis (qOSMA). The assay is based on the combination of methylation-sensitive FastDigest® endonuclease digestion and real-time quantitative PCR (qPCR) in a single reaction, with reaction conditions providing DNA digestion in a first step, followed by endonuclease inactivation and SybrGreen-based qPCR. The degree of DNA methylation is determined by comparing the quantification cycles (Cq) of reactions containing either a methylation-sensitive endonuclease or a sham mixture with no endonuclease. Control reactions interrogating an unmethylated locus detect possible endonuclease inhibitors and can be used for simultaneous copy number assessment by means of standard qPCR with the ΔΔCq method. As proof of principle, we diagnosed the neurogenetic disorders Prader-Willi syndrome (PWS) and Angelman syndrome (AS) using the qOSMA assay. PWS and AS evolve from deletions or uniparental disomies spanning the imprinting locus SNRPN, resulting in SNRPN methylation levels of 100% (PWS), 0% (AS) or 50% (healthy individuals). We correctly diagnosed PWS and AS in 35 individuals, with SNRPN methylation measurements being in the range of 92.0 – 103.5% (PWS, n = 12), 1.0 – 6.7% (AS, n = 12) and 48.3 – 60.1% (healthy controls, n = 11). PWS and AS patients carrying a deletion were simultaneously identified by applying the ΔΔCq method to sham reactions, resulting in ratios of 0.40 – 0.62 (deleted, n = 19) and 0.87 – 1.09 (non-deleted, n = 16). In addition, we analyzed mixes of PWS and AS DNA to assess whether qOSMA is truly quantitative for all methylation levels. Here, the methylation measurements did not deviate more than 5.9% from the expected values, resulting in a R2 of 0.997. We finally found that DNA quality was not an issue with samples (re-)extracted with silica-based protocols, yet we encountered incomplete digestion with phenol-chloroform extracted and old DNA samples. To correct for such reduced DNA digestion, we developed an algorithm which implements DNA digestibility information obtained with the control reactions of the unmethylated locus, allowing application of qOSMA to low-quality DNA. Such correction is of particular interest for retrospective studies investigating DNA of variable quality. In summary, we present the first closed-tube one-step assay for quantitative methylation analysis using native genomic DNA as template. The assay minimizes hands-on time and is accomplished within 1.5 h after DNA extraction, predestining it for high-throughput methylation analyses.