Longitudinal epigenetic and gene expression profiles analyzed by three-component analysis reveal down-regulation of genes involved in protein translation in human aging

Abstract

Data on biological mechanisms of aging are mostly obtained from cross-sectional study designs. An inherent disadvantage of this design is that inter-individual differences can mask small but biologically significant age-dependent changes. A serially sampled design (same individual at different time points) would overcome this problem but is often limited by the relatively small numbers of available paired samples and the statistics being used. To overcome these limitations, we have developed a new vector-based approach, termed three-component analysis, which incorporates temporal distance, signal intensity and variance into one single score for gene ranking and is combined with gene set enrichment analysis. We tested our method on a unique age-based sample set of human skin fibroblasts and combined genome-wide transcription, DNA methylation and histone methylation (H3K4me3 and H3K27me3) data. Importantly, our method can now for the first time demonstrate a clear age-dependent decrease in expression of genes coding for proteins involved in translation and ribosome function. Using analogies with data from lower organisms, we propose a model where age-dependent down-regulation of protein translation-related components contributes to extend human lifespan.

Figure 1.

Schematics of the GSEA-guided 3CA analysis framework. (A) Illustration of the experimental design used for the study. (B) Overview of the experimental and data analysis steps.

Figure 2.

Graphical representation of the 3CA approach. Individual sample pairs can be visualized as dots in a plane. Significant age-related changes can be identified as an optimal intersection of signal intensity, sample variance and temporal change.

Figure 3.

3CA-guided GSEA detecting the most significant gene sets correlated with temporal up- or down-regulation. Temporal expression and epigenetic changes can be evaluated in one plot. For a comprehensive view, we plotted the 3CA score distributions for gene expression (RNA-seq), DNA methylation (meDNA), H3K4-trimethylation (H3K4me3) and H3K27-trimethylation (H3K27me3) for the specific gene sets (set) and for all genes (total) (left panels). In addition to the box charts, we plotted a scattered distribution of the scores (right panels). In this example, we examined the significant 3CA scores for gene expression for matched promoter regions of the epigenetic sets. In order to access the significance for up- or down-regulation, we obtained the temporal changes for the specific gene sets and performed GSEA with the obtained pre-ranked list (40). As we pre-filtered for significant 3CA scores, the respective distribution of temporal changes will be pushed to the edges of the GSEA plots. The values at the y-axis of the enrichment plots are the maximum and minimum enrichment score values obtained from the temporal changes. Shown are all identified processes connected with protein turnover. A: translation, B: cytosolic ribosome, C: proteasome, D: chaperone, E: lysosome, F: protein transport, G: ubiquitination.

Figure 4.

Venn diagrams showing overlap between epigenetic and transcriptional data sets for the top 3CA scores. Transcripts which were preferentially up-regulated with age were compared with down-regulated H3K27me3 regions, down-regulated DNA methylation regions, and up-regulated H3K4me3 regions and vice versa (panels A and B, respectively). Data included are within 1 SD from the best score (top ∼2%).

Figure 5.

Functional enrichment analysis for CGIs. Data are for the top scored CGIs within 1 SD from the best score for both up- and down-regulated regions. Functional significance of top scored CGIs for age-related changes in H3K27me3 (A) and DNA methylation (B), defined by GREAT analysis and using terms of the Molecular Signature Database (MsigDB).