Milea Timbergen

94 The findings of Song et al. prompted us to hypothesize that the different mutations found in CTNNB1 in DTF affect interacting DNMT1 differently consequently causing altered DNA methylation patterns. This study investigates DNA methylation patterns of the two most common mutation types of DTF (S45F and T41A), aiming to provide insight in the biological underpinnings of the different clinical behaviour of these DTF mutants. Materials and Methods Patient and sample selection Patients with histologically proven, primary DTF and a S45F or T41A CTNNB1 mutation were identified in the Erasmus MC pathology database. Corresponding formalin-fixed paraffin-embedded (FFPE) tumour tissue blocks were collected from the Erasmus MC tissue bank. Similarly, clinicopathological characteristics such as sex, age at diagnosis, tumour site (extra-abdominal, intra-abdominal, or abdominal wall), and largest tumour size (in millimetres [mm]) on imaging were obtained from the patient files. The CTNNB1 exon 3 mutations were previously determined for diagnostic purposes essentially as described by Dubbink et al. 27 . In short: Tumour DNA was extracted from FFPE tumour tissue using proteinase K and 5% Chelex®-100 chelating resin (Bio-Rad). Sequence analysis of CTNNB1 exon 3 was performed by bidirectional sequencing of PCR-amplified DNA fragments using M13-tailed forward and reverse primers. The selected patients did not receive any treatment before the specimens were obtained. An expert soft tissue sarcoma pathologist confirmed the diagnosis by examining hematoxylin-eosin stained sections of the FFPE samples. DNA isolation DNA was isolated from five consecutive FFPE DTF sections of 10µm using the Allprep DNA/ RNA kit according to the manufacturer’s recommendations (Qiagen, Hilden, Germany). The DNA quality and the concentrations were determined using a Nanodrop-2000 (Isogen Life Science, Utrecht, and the Netherlands). The 260nm/280nm ratio was ≥ 1.80 for all DNA preparations. MeD-seq sample preparations Methylated DNA sequencing (MeD-seq) was used to analyse genome wide DNA methylation. MeD-seqprovides single-nucleotide resolution by exploiting the properties of the DNA methylation dependent restriction enzyme LpnPI 28 . This enzyme generates DNA 4