Abstract
Aberrant DNA methylation is one of the most common alterations in cancer, and a vast diversity of methods for its investigation exists. Quantitative methylation-specific PCR (qMSP) is frequently used to estimate the amounts of methylation at specific loci, such as gene promoters. However, diverging qMSP results are being reported in the literature, underscoring the need for standardization of the individual steps of the protocol. In this study we aimed at identifying the most likely sources of variability in the qMSP method by investigating six individual gene promoters. Data from over 150 rounds of qMSP confirmed that the choice of normalization reference is crucial. Use of the repetitive element ALU results in remarkably less divergence than does use of the single/low-copy genes ACTB and COL2A1. Importantly, a deletion or amplification of such genes in cancer cells may cause a doubling or halving of the PMR values, respectively. Furthermore, careful control of the DNA input amount is essential, both in the bisulfite conversion reaction, but more importantly in the subsequent qMSP, as normalization by the reference control is only successful within a limited range of template variation. Additionally, storage of bisulfite-treated DNA could cause some PMR variation, and direct qMSP analysis of samples is therefore recommended. Importantly, up to 20% variation in PMR values should be expected, and taken into account when presenting differential methylation. A standardized qMSP pipeline has been suggested, optimized to give the most consistent PMR results. Knowledge about DNA promoter hypermethylation in malignant peripheral nerve sheet tumor (MPNST), a rare cancer of the nervous system, is limited compared to other cancer types. Thus, we intended to identify novel MPNST promoter methylation candidates. Genes significantly upregulated in MPNST cell lines after treatment with the demethylating agents AZA/TSA, and downregulated in malignant- compared to benign tissue, were chosen for MSP analysis. The two best-performing candidates were further subjected to bisulfite sequencing, and finally analyzed by qMSP in a large MPNST patient series (n=91). This resulted in identification of HCK and SPINT2 with 34% and 42% promoter methylation, respectively. Both loci were cancer specifically methylated, in the sense that all controls (n=22) were unmethylated.