Detection and evaluation of DNA methylation markers found at SCGN and KLF14 loci to estimate human age

•Single- and dual-locus age-predictors for saliva were developed from CpG sites in KLF14 and SCGN using a multivariate linear regression analysis.•The single-locus model explained 85% of age variance with a MAD of 5.8 years and the dual-locus model explained 84% of age variance with a MAD of 6.2 years in the training set.•The single-locus model predicted age (±8 years) with accuracy of 78.9% for younger subjects (≤40 years) and a MAD of 5.1 years in the validation set.•The age predictor for blood was based on CpG sites in SCGN and KLF14 and accounted for 71% of age variance with a MAD of 6.6 years in the training set.

Recent developments in the analysis of epigenetic DNA methylation patterns have demonstrated that certain genetic loci show a linear correlation with chronological age. It is the goal of this study to identify a new set of epigenetic methylation markers for the forensic estimation of human age. A total number of 27 CpG sites at three genetic loci, SCGN, DLX5 and KLF14, were examined to evaluate the correlation of their methylation status with age. These sites were evaluated using 72 blood samples and 91 saliva samples collected from volunteers with ages ranging from 5 to 73 years. DNA was bisulfite modified followed by PCR amplification and pyrosequencing to determine the level of DNA methylation at each CpG site. In this study, certain CpG sites in SCGN and KLF14 loci showed methylation levels that were correlated with chronological age, however, the tested CpG sites in DLX5 did not show a correlation with age.Using a 52-saliva sample training set, two age-predictor models were developed by means of a multivariate linear regression analysis for age prediction. The two models performed similarly with a single-locus model explaining 85% of the age variance at a mean absolute deviation of 5.8 years and a dual-locus model explaining 84% of the age variance with a mean absolute deviation of 6.2 years. In the validation set, the mean absolute deviation was measured to be 8.0 years and 7.1 years for the single- and dual-locus model, respectively. Another age predictor model was also developed using a 40-blood sample training set that accounted for 71% of the age variance. This model gave a mean absolute deviation of 6.6 years for the training set and 10.3 years for the validation set. The results indicate that specific CpGs in SCGN and KLF14 can be used as potential epigenetic markers to estimate age using saliva and blood specimens. These epigenetic markers could provide important information in cases where the determination of a suspect's age is critical in developing investigative leads.