Tumor samples
Tumor samples were obtained during surgery from 317 primary HNSCC tissues and 225
normal tissues. All patients were treated at the Department of Otolaryngology, Hamamatsu University School of Medicine. All patients provided written informed consent under a protocol approved by our Institutional Review Board (Approval Date: October 2, 2015; Ethics Approval Code: 25-149). The average age of the patients was 65.8 years (range: 32–92 years). Among the patients, 261 were men and 56 were women. The primary tumor locations were the larynx (n = 64), hypopharynx (n = 87), oropharynx (n = 68), and oral cavity (n = 98). Clinical information was obtained from the clinical records.
Bisulfite modification and quantitative real-time methylation and unmethylation PCR
Genomic DNA was extracted from fresh tissues using the QIAamp DNA Mini Kit (QIAGEN, Hilden, Germany). Sodium bisulfite conversion was performed using the MethylEasy Xceed Rapid DNA Bisulfite Modification Kit (TaKaRa, Tokyo, Japan). The primer sequences specific for LINE-1 and TSGs are shown in Table S1 (Additional file 1: Table S1) (14). A standard curve for quantitative real-time methylation and unmethylation PCR (Q-MSP and Q-UMSP) was constructed by plotting five serially diluted standard solutions of EpiScope Methylated HCT116 gDNA (TaKaRa, Tokyo, Japan) and EpiScope® Unmethylated HCT116 DKO gDNA (TaKaRa, Tokyo, Japan). The normalized methylation value (NMV) was defined as follows: NMV = (TSGs-S/TSGs-FM)/(ACTB-S/ACTB-FM), where TSGs-S and TSGs-FM represent TSGs methylation level in sample and universally methylated DNA, respectively, and ACTB-S and ACTB-FM correspond to b-actin levels in samples and universally methylated DNA, respectively (15). Known amounts of LINE-1 methylated and unmethylated DNA molecules were used to generate absolute standard curves (y = −6.125 * log(x) + 26.54 and y = −2.765 * log(x) + 21.19, respectively). The number of methylated or unmethylated LINE-1 sequence copies was extrapolated from the standard curves. The percentage of methylation was defined as the ratio between methylated molecules and the sum of methylated and unmethylated molecules [unmethylated copy number/ (methylated copy number + unmethylated copy number)].
Enzyme-linked immunosorbent assay for 5-hmC quantification
The 5-hmC content of genomic DNA was determined using a Quest 5-hmC DNA enzyme-linked immunosorbent assay (ELISA) Kit (Zymo Research, Irvine, CA, USA). The assays were performed according to the manufacturer’s instructions, loading 100 ng of DNA per well and using 4 μg/mL of anti-5hmC polyclonal antibodies. Absorbance at 430 nm was measured using a SynergyH1 microplate reader and Gen5 software (BioTek, Winooski, VT, USA). The amount of 5-hmC was calculated as a percentage based on a standard curve generated using kit controls (16).
Liquid biopsy
We tested LINE-1 hypomethylation levels in a validation study comprising five OCC patients using ctDNA obtained pre- and post-treatment. We isolated ctDNA from 4.0 mL plasma samples using affinity-based binding to magnetic beads as per manufacturer’s instructions (QIAamp MinElute ccfDNA Kit, QIAGEN, Hilden, Germany). Peripheral blood samples (10 mL each) were collected in cell-stabilizing tubes (Cell-Free DNA Collection Tube, Roche, CA, USA).
Data and Statistical Analyses
The LINE-1 hypomethylation levels in 317 tumor samples as well as the overall patient characteristics were analyzed statistically. Receiver-operator characteristic (ROC) curve analyses were performed in 225 matched pair samples to compare LINE-1 hypomethylation levels between tumor and normal tissues. Kaplan-Meier curves and log-rank tests were used to estimate DFS in the different subgroups. Differences in the LINE-1 hypomethylation levels among the 317 patients according to their clinical information were examined using Fisher's exact test. The prognostic value of methylation status was assessed using multivariate Cox proportional hazards analysis adjusting for age (≥ 75 versus < 75 years), sex, smoking status, alcohol intake, and tumor stage (I, II, and III versus IV). Spearman's correlation analysis was used to identify whether 5-hmC levels or TSG methylation indexes (MI) affect LINE-1 hypomethylation. P values < 0.05 were considered statistically significant. Statistical analyses were performed using StatMate IV software (ATMS Co. Ltd., Tokyo, Japan) and the Stata/SE 13.0 system (Stata Corporation, TX, USA).