DNA methylation age estimation from human bone and teeth

Bones and teeth are valuable sources of information in forensic contexts. However, few studies have considered these tissues for DNA methylation (DNAm) analyses in age estimation. The main goal of the present study was to develop DNAm-based age prediction models (APMs) specific for bones and teeth. DNA samples from Portuguese individuals were evaluated through bisulphite Sanger sequencing and a multiplex SNaPshot assay to analyse relationships between DNAm and age. In bones our results allowed to build a final APM through Sanger sequencing with six CpGs at genes ELOVL2, EDARADD and MIR29B2C, explaining 92.5% of age variation, with a mean absolute deviation (MAD) from chronological age of 2.56 years. For teeth, we were unable to build a final multi-locus APM; the best FHL2 CpG4 age predictor explained 41.3% of age variance with a MAD of 11.35 years. The SNaPshot assay allowed to build two final dual-locus APMs with CpGs at FHL2 and KLF14 for bones, explaining 57.6% of age variation (MAD of 7.18 years), and with CpGs at ELOVL2 and KLF14 for teeth, explaining 76.4% of age variation (MAD of 7.07 years). Our study showed the usefulness of bone and tooth samples in evaluation of DNAm levels for age estimation purposes.