Mappatura della 5-formilcitosina a livello di singolo nucleotidenella mucosa del colon sano usando tecniche di sequenziamento basate sul bisolfito

DNA methylation is a well-characterized epigenetic mark involving the transfer of a methyl group to C-5 position of the cytosine ring. In mammals, it usually occurs in a CpG dinucleotide context and is able to influence many processes as genomic stability, gene expression and development. Recently, several DNA demethylation mechanisms have been proposed and, in particular, a possible role for Ten Eleven Translocator (TET) proteins in 5-methylcytosine (5mC) oxidation has emerged. TET enzymes are able to convert 5mC into 5-hydroxymethylcytosine (5hmC), which can be further oxidized into 5-formylcytosine (5fC) and 5-carboxylcytosine (5caC). These two bases can be efficiently excised by thymine DNA glycosylase (TDG) and restored to unmodified cytosine by base excision repair pathway. Besides being intermediates of the demethylation process, oxidized cytosines may also fulfill biological roles. Presence of 5mC oxidized derivatives at particular genomic locations, holding regulatory functions, and identification of specific reader proteins for these modifications, as transcription factors and chromatin modifying enzymes, has strengthened this hypothesis. In addition, alterations of oxidized cytosine levels have been observed in many types of cancer, suggesting an involvement of these nucleotides in tumor development. Global depletion of 5hmC, 5caC and, in particular, 5fC has been reported in colon carcinoma too. The aim of the present study was therefore to investigate 5fC presence in colon mucosa in order to analyze its amount and genomic localization. Because of their extremely low abundance, charting oxidized bases is technically challenging. Single-nucleotide resolution methods, based on bisulfite conversion of the DNA and sequencing, were applied in order to map 5fC in the genome. In particular, formyl-chemically assisted bisulfite (fCAB) treatment, coupled with reduced representation sequencing (RRBS), was used and allowed the detection of a high number of CpG sites marked by 5fC. Validation of the data was performed by deep-sequencing of PCR products (amplicons) generated on specific genomic regions showing higher level of 5-formylcytosine. The significant amount of 5fC detected by fCAB-RRBS was not confirmed by this analysis. Applying DNA digestion before fCAB treatment, seemed, however, to improve 5-formylcytosine detection in amplicons, leading to results more concordant with fCAB-RRBS data, at least in some regions. This could indicate that the treatment may be influenced by DNA size and secondary structure. Methylase-assisted bisulfite sequencing (MAB-Seq) was also performed to further analyzed 5fC distribution in the genome. Only a small number of CpG sites showing 5fC presence appeared to be in common by comparing both genome-wide methods. In addition, regions of 5fC accumulation identified by fCAB-RRBS, and selected for amplicon analysis, were mainly depleted for oxidized bases when analyzed by MAB-Seq. The data would therefore point to a variability in 5fC detection when different approaches are applied, indicating that technical issues associated to each method could prevent a correct profiling of 5fC in the genome. Validation of the obtained results by deep-sequencing, careful data analysis and improvement of mapping strategies are therefore needed to allow an accurate and reliable identification of 5-formylcytosine and evaluate its possible functional roles.