Research Paper Volume 15, Issue 14 pp 6690—6709

Parallel shift of DNA methylation and gene expression toward the mean in mouse spleen with aging

Sangkyun Jeong1,2, , Sunwha Cho2, , Seung Kyoung Yang2, , Soo A. Oh1, , Yong-Kook Kang3, ,

  • 1 Medical Research Division, Korea Institute of Oriental Medicine (KIOM), Yuseong-gu, Daejeon 34054, South Korea
  • 2 Genomics Department, Keyomics Co. Ltd., Yuseong-gu, Daejeon 34013, South Korea
  • 3 Development and Differentiation Research Center, Aging Convergence Research Center (ACRC), Korea Research Institute of Bioscience Biotechnology (KRIBB), Yuseong-gu, Daejeon 34141, South Korea

Received: April 19, 2023       Accepted: July 6, 2023       Published: July 25, 2023      

https://doi.org/10.18632/aging.204903
How to Cite

Copyright: © 2023 Jeong et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Abstract

Age-associated DNA-methylation drift (AMD) manifests itself in two ways in mammals: global decrease (hypomethylation) and local increase of DNA methylation (hypermethylation). To comprehend the principle behind this bidirectional AMD, we studied methylation states of spatially clustered CpG dinucleotides in mouse splenic DNA using reduced-representation-bisulfite-sequencing (RRBS). The mean methylation levels of whole CpGs declined with age. Promoter-resident CpGs, generally weakly methylated (<5%) in young mice, became hypermethylated in old mice, whereas CpGs in gene-body and intergenic regions, initially moderately (~33%) and extensively (>80%) methylated, respectively, were hypomethylated in the old. Chromosome-wise analysis of methylation revealed that inter-individual heterogeneities increase with age. The density of nearby CpGs was used to classify individual CpGs, which found hypermethylation in CpG-rich regions and hypomethylation in CpG-poor regions. When genomic regions were grouped by methylation level, high-methylation regions tended to become hypomethylated whereas low-methylation regions tended to become hypermethylated, regardless of genomic structure/function. Data analysis revealed that while methylation level and CpG density were interdependent, methylation level was a better predictor of the AMD pattern representing a shift toward the mean. Further analysis of gene-expression data showed a decrease in the expression of highly-expressed genes and an increase in the expression of lowly-expressed genes with age. This shift towards the mean in gene-expression changes was correlated with that of methylation changes, indicating a potential link between the two age-associated changes. Our findings suggest that age-associated hyper- and hypomethylation events are stochastic and attributed to malfunctioning intrinsic mechanisms for methylation maintenance in low- and high-methylation regions, respectively.

Abbreviations

CpG: CpG dinucleotide; CGI: CpG island; RRBS: reduced-representation bisulfite sequencing; AMD: age-associated DNA methylation drift; SiNG-PCRseq: spiking-in a neighbor genome for competitive PCR amplicon sequencing; CV: coefficient of variation.