Identifying 5-hydroxymethylcytosine as a potential cancer biomarker using FFPE DNA samples

Head and neck cancer (HNC) is the seventh most common cancer worldwide and identifying biomarkers for its diagnosis, prognosis, and treatment success monitoring is crucial. Recent studies suggested that 5-hydroxymethylcytosine (5hmC), a modified DNA base that can impact gene expressions, has great potential as a biomarker for the diagnosis of various cancers. However, whether 5hmC can be used as a biomarker for HNC detection remains unexplored. In contrast to fresh HNC tissues from patients, formalin-fixed and paraffin-embedded (FFPE) samples are much more readily available for biomarker seeking. However, FFPE samples usually have severe DNA damage, leading to a large percentage of ssDNA fragments and making the sample incompatible with the current technology for 5hmC profiling. We hypothesized that using a new approach for 5hmC profiling that is compatible with ssDNA fragments would allow us to identify the potential 5hmC biomarker for HNC diagnosis in FFPE samples by comparing the differences in 5hmC distribution between tumor and adjacent normal tissues. Here, we report the use of an improved CMS-seq. method to sequence FFPE samples from HNC tumors and their adjacent normal tissues to generate pairs of genomic data from four different patients. After sequencing and data analysis, we identified 339 genes with differentiable 5hmC levels (p-value <0.05). Among them, three genes (PRKD2, HADHA, and AIPL1) have p-adjusted values less than 0.05, suggesting that the distinct 5hmC pattern in these three genes has promising potential as biomarkers for HNC diagnosis.