Development of epigenetic aging clocks by novel targeted sequencing approaches

Yang, Han; Wagner, Wolfgang (Thesis advisor); Zenke, Martin (Thesis advisor)

Aachen : RWTH Aachen University (2022)
Dissertation / PhD Thesis

Dissertation, RWTH Aachen University, 2022

Abstract

Age-associated DNA methylation (DNAm) is highly reproducible between individuals, providing a promising biomarker of aging, namely ‘epigenetic clock’. While previous epigenetic clocks based on genome-wide DNAm profiles enable robust age predictions by numerous CG dinucleotides (CpGs), targeted DNAm assays focused on only a few CpGs provide better alternatives with higher efficiency and easier standardization. In this thesis, we measured DNAm of new sets of age-associated CpGs from genome- wide profiles and then developed epigenetic clocks in both human and mice using several targeted approaches. Although all of these methods offer similar robust age prediction accuracy as genome-wide clocks, pyrosequencing, and droplet digital PCR (ddPCR) seem to be more accurate. In particular, ddPCR could largely reduce the amplification bias of bisulfite converted DNA sequences and thus could be a promising method for site-specific DNAm analysis. Barcoded bisulfite amplicon sequencing (BBA-seq) detects DNAm within multiple amplicons in parallel with single-strand resolution. Interestingly, we found that neighboring age-associated CpGs seem to be regulated in a stochastic manner. Further correlation analysis of DNAm levels with age at neighboring CpG sites reveals a bell-shaped curve, which often associated with a CCCTC-binding factor (CTCF) binding site. We also demonstrated that single BBA- seq reads are capable to predict donor age and thus reflected the heterogeneity of age-associated DNAm. Numerous studies have confirmed that nutritional interventions can reduce biological age and slow down aging. However, we observed no significant changes in metabolic health or epigenetic clock in mice with 12 weeks of fat and protein diets, except that high fat diets led to a clear weight gain. This reveals that short-term dietary interventions may not be sufficient to alter the metabolic and aging status, and thus prolonged interventions may be indispensable to achieve significant treatment effects. Taken together, we optimized different epigenetic clocks by novel targeted sequencing approaches and demonstrated their robust age predictions for humans and mice, which may better support standardized and cost-effective high-throughput epigenetic age prediction in forensic and diagnostic applications. Furthermore, these clocks should also provide better insights into deciphering the mechanisms of the biology of aging.

Institutions

• Department of Biology [160000]
• [811002-3]