Age estimation in forensic anthropology remains a challenge, particularly in adults, where accuracy tends to decrease with age. As a key component of the biological profile, improving age estimation methods is essential for human identification. DNA methylation has recently emerged as a promising tool to complement traditional anthropological approaches. Skeletal tissues, such as bones and teeth, are often the only samples available in advanced stages of decomposition. Teeth are especially valuable due to their low porosity, high mineral content, and anatomical protection, thereby enhancing the preservation of biomolecules.
Despite their potential, studies on dental tissues remain limited, with most research focusing on molars and whole teeth, and only a few studies assessing differences among tissue types (e.g., pulp, dentin, cementum). However, variability related to tooth type has not been thoroughly explored. This study investigates whether methylation patterns differ between incisors and molars within the same tissue (dentin), and how such differences may influence age‑estimation models.
Dentin was isolated from 15 molars and 11 incisors (aged 16–88 years). DNA was extracted using the DNeasy® Blood and Tissue Kit (Qiagen), quantified using the Qubit™ HS dsDNA Assay (Thermo Fisher Scientific), and 200 ng were bisulfite-converted using the EZ DNA Methylation-Lightning™ Kit (Zymo Research). PCR amplification was performed using the PyroMark PCR Master Mix (Qiagen), and pyrosequencing was carried out on the PyroMark® Q48 Autoprep system (Qiagen). Multiple CpG sites from five previously studied age‑associated genes (ELOVL2, FHL2, NPTX2, KLF14, and SCGN) were analyzed. Statistical modeling was conducted using stepwise linear regression in SPSS.
Distinct prediction models were obtained for each tooth type. In incisors, the prediction model achieved an MAE of 2.8 years (SD = 2.2), R² = 0.98, with major contributors from SCGN (CpG2), NPTX2 (CpG7, CpG14), and KLF14 (CpG5). In contrast, the molar-based model included only CpGs from NPTX2 (CpG6, CpG10, CpG12), yielding an MAE of 6.06 years (SD = 4.69) and R² = 0.795.
These results indicate that different CpG markers may drive age prediction depending on tooth type, even within the same tissue. Differences in developmental timing—incisors forming earlier in childhood and molars later in adolescence—along with structural, metabolic, and environmental variation, may contribute to distinct methylation signatures. This study provides initial evidence of intratissue variability in epigenetic profiles associated with tooth type and highlights the need for tooth-specific models to improve the accuracy of epigenetic age estimation in forensic applications.