Age Prediction of Human Based on DNA Methylation by Blood Tissues

Male-specific age prediction based on Y-chromosome DNA methylation with blood using pyrosequencing

Age prediction is an important aspect of forensic science that offers valuable insight into identification. In recent years, extensive studies have been conducted on age prediction based on DNA methylation, and numerous studies have demonstrated that DNA methylation is a reliable biomarker for age prediction. However, almost all studies on age prediction based on DNA methylation have focused on age-related CpG sites in autosomes, which are concentrated on single-source DNA samples. Mixed samples, especially male-female mixed samples, are common in forensic casework. The application of Y-STRs and Y-SNPs can provide clues for the genetic typing of male individuals in male-female mixtures, but they cannot provide the age information of male individuals. Studies on Y-chromosome DNA methylation can address this issue. In this study, we identified five age-related CpG sites on the Y chromosome (Y-CpGs) and developed a male-specific age prediction model using pyrosequencing combined with a support vector machine algorithm. The mean absolute deviation of the model was 5.50 years in the training set and 6.74 years in the testing set. When we used a male blood sample to predict age, the deviation between the predicted and chronological age was 1.18 years. Then, we mixed the genomic DNA of the male and a female at ratios of 1:1, 1:5, 1:10, and 1:50, the range of deviation between the predicted and chronological age of the male in the mixture was 1.16-1.74 years. In addition, there was no significant difference between the methylation values of bloodstains and blood in the same sample, which indicates that our model is also suitable for bloodstain samples. Overall, our results show that age prediction using DNA methylation of the Y chromosome has potential applications in forensic science and can be of great help in predicting the age of males in male-female mixtures. Furthermore, this work lays the foundation for future research on age-related applications of Y-CpGs.

Epigenetic-based age prediction in blood samples: Model development

Aging is a complex process influenced by genetic, epigenetic, and environmental factors that lead to tissue deterioration and frailty. Epigenetic mechanisms, such as DNA methylation, play a significant role in gene expression regulation and aging. This study presents a new age estimation model developed for the Turkish population using blood samples. Eight CpG sites in loci TOM1L1, ELOVL2, ASPA, FHL2, C1orf132, CCDC102B, cg07082267, and RASSF5 were selected based on their correlation with age. Methylation patterns of these sites were analyzed in blood samples from 100 volunteers, grouped into age categories (20-35, 36-55, and ≥56). Sensitivity analysis indicated a reliable performance with DNA inputs ≥1 ng. Statistical modeling, utilizing Multiple Linear Regression, underscores the reliability of the primary 6-CpG model, excluding cg07082267 and TOM1L1. This model demonstrates strong correlations with chronological age (r = 0.941) and explains 88% of the age variance with low error rates (MAE = 4.07, RMSE = 5.73 years). Validation procedures, including a training-test split and fivefold cross-validation, consistently confirm the model's accuracy and consistency. The study indicates minimal variation in error scores across age cohorts and no significant gender differences. The developed model showed strong predictive accuracy, with the ability to estimate age within certain prediction intervals. This study contributes to the age prediction by using DNA methylation patterns, which can have disparate applications, including forensic and clinical assessments.

Building Minimized Epigenetic Clock by iPlex MassARRAY Platform

Epigenetic clocks are valuable tools for estimating both chronological and biological age by assessing DNA methylation levels at specific CpG dinucleotides. While conventional epigenetic clocks rely on genome-wide methylation data, targeted approaches offer a more efficient alternative. In this study, we explored the feasibility of constructing a minimized epigenetic clock utilizing data acquired through the iPlex MassARRAY technology. The study enrolled a cohort of relatively healthy individuals, and their methylation levels of eight specific CpG dinucleotides in genes SLC12A5, LDB2, FIGN, ACSS3, FHL2, and EPHX3 were evaluated using the iPlex MassARRAY system and the Illumina EPIC array. The methylation level of five studied CpG sites demonstrated significant correlations with chronological age and an acceptable convergence of data obtained by the iPlex MassARRAY and Illumina EPIC array. At the same time, the methylation level of three CpG sites showed a weak relationship with age and exhibited a low concordance between the data obtained from the two technologies. The construction of the epigenetic clock involved the utilization of different machine-learning models, including linear models, deep neural networks (DNN), and gradient-boosted decision trees (GBDT). The results obtained from these models were compared with each other and with the outcomes generated by other well-established epigenetic clocks. In our study, the TabNet architecture (deep tabular data learning architecture) exhibited the best performance (best MAE = 5.99). Although our minimized epigenetic clock yielded slightly higher age prediction errors compared to other epigenetic clocks, it still represents a viable alternative to the genome-wide epigenotyping array.

Developing a male-specific age predictive model based on Y-CpGs for forensic analysis

In forensic work, predicting the age of the criminal suspect or victim could provide beneficial clues for investigation. Epigenetic age estimation based on age-correlated DNA methylation has been one of the most widely studied methods of age estimation. However, almost all available epigenetic age prediction models are based on autosomal CpGs, which are only applicable to single-source DNA samples. In this study, we screened the available methylation data sets to identify loci with potential to meet the objectives of this study and then established a male-specific age prediction model based on 2 SNaPshot systems that contain 13 Y-CpGs and the mean absolute deviation (MAD) values were 4-6 years. The multiplex methylation SNaPshot systems and age-predictive model have been validated for sensitivity (the DNA input could be as low as 0.5 ng) and male specificity. They are supposed to have feasibility in forensic practice. In addition, it demonstrated that the method was also applicable to bloodstains, which were commonly found at crime scenes. The results showed good performance (the training set: R² = 0.9341, MAD = 4.65 years; the test set: R² = 0.8952, MAD = 5.73 years) in case investigation for predicting male age. For mixtures, when the male to female DNA ratio is 1:1, 1:10, the deviation between the actual age and the predicted age obtained by the model was less than 8 years, which offers great hope for future prediction of the age of males in mixtures and will be a powerful tool for special cases, such as sexual assault. Furthermore, the work provides a basis for the application of Y-CpGs in forensic science.

DNA methylation of the TPMT gene and azathioprine pharmacokinetics in children with very early onset inflammatory bowel disease

BACKGROUND

Thiopurine methyltransferase (TPMT) is a crucial enzyme for azathioprine biotransformation and its activity is higher in very early onset inflammatory bowel disease (VEO-IBD) patients than in adolescents with IBD (aIBD).

AIMS

The aims of this pharmacoepigenetic study were to evaluate differences in peripheral blood DNA methylation of the TPMT gene and in azathioprine pharmacokinetics in patients with VEO-IBD compared to aIBD.

METHODS

The association of age with whole genome DNA methylation profile was evaluated in a pilot group of patients and confirmed by a meta-analysis on 3 cohorts of patients available on the public functional genomics data repository. Effects of candidate CpG sites in the TPMT gene were validated in a larger cohort using pyrosequencing. TPMT activity and azathioprine metabolites (TGN) were measured in patients' erythrocytes by HPLC and associated with patients' age group and TPMT DNA methylation.

RESULTS

Whole genome DNA methylation pilot analysis, combined with the meta-analysis revealed cg22736354, located on TPMT downstream neighboring region, as the only statistically significant CpG whose methylation increases with age, resulting lower in VEO-IBD patients compared to aIBD (median 9.6% vs 12%, p = 0.029). Pyrosequencing confirmed lower cg22736354 methylation in VEO-IBD patients (median 4.0% vs 6.0%, p = 4.6 ×10^-5). No differences in TPMT promoter methylation were found. Reduced cg22736354 methylation was associated with lower TGN concentrations (rho = 0.31, p = 0.01) in patients with VEO-IBD and aIBD.

CONCLUSION

Methylation of cg22736354 in TPMT gene neighborhood is lower in patients with VEO-IBD and is associated with reduced azathioprine inactivation and increased TGN concentrations.

How Microbiomes Affect Skin Aging: The Updated Evidence and Current Perspectives

The skin has a multifactorial aging process, caused by both intrinsic and extrinsic factors. A major theory of aging involves cellular senescence or apoptosis resulting from oxidative damage as the skin’s antioxidant system tends to weaken with age. The human microbiota is a complex ecosystem that is made up of microorganisms (bacteria, fungi, and viruses). Both gut and skin microbiota have essential roles in the protection against invading pathogens, mediating inflammatory conditions, and the modulation of the immune system which is involved in both innate and adaptive immune responses. However, the human microbiome could be changed during the life stage and affected by various perturbations. An alteration of the intestinal bacteria results in “microbial dysbiosis” which is associated with the influence of various diseases, including aging. The skin interactome is a novel integration of the “genome-microbiome-exposome” that plays a significant role in skin aging and skin health. Mitigating the negative impacts of factors influencing the skin interactome should be the future strategy to protect, prevent, and delay skin aging along with preserving healthy skin conditions. This review summarizes the current evidence on how human microbiomes affect skin aging and demonstrates the possible interventions, relating to human microbiomes, to modulate skin health and aging. Probiotics-based products are currently available mainly for the add-on treatment of many dermatologic conditions. However, at this point, there are limited clinical studies on skin anti-aging purposes and more are required as this evolving concept is on the rise and might provide an insight into future therapeutic options.