Transcriptome analysis reveals the difference between "healthy" and "common" aging and their connection with age-related diseases

Abstract

A key goal of aging research was to understand mechanisms underlying healthy aging and develop methods to promote the human healthspan. One approach is to identify gene regulations unique to healthy aging compared with aging in the general population (i.e., "common" aging). Here, we leveraged Genotype-Tissue Expression (GTEx) project data to investigate "healthy" and "common" aging gene expression regulations at a tissue level in humans and their interconnection with diseases. Using GTEx donors' disease annotations, we defined a "healthy" aging cohort for each tissue. We then compared the age-associated genes derived from this cohort with age-associated genes from the "common" aging cohort which included all GTEx donors; we also compared the "healthy" and "common" aging gene expressions with various disease-associated gene expressions to elucidate the relationships among "healthy," "common" aging and disease. Our analyses showed that 1. GTEx "healthy" and "common" aging shared a large number of gene regulations; 2. Despite the substantial commonality, "healthy" and "common" aging genes also showed distinct function enrichment, and "common" aging genes had a higher enrichment for disease genes; 3. Disease-associated gene regulations were overall different from aging gene regulations. However, for genes regulated by both, their regulation directions were largely consistent, implying some aging processes could increase the susceptibility to disease development; and 4. Possible protective mechanisms were associated with some "healthy" aging gene regulations. In summary, our work highlights several unique features of GTEx "healthy" aging program. This new knowledge could potentially be used to develop interventions to promote the human healthspan.

Keywords: GTEx; age-related diseases; common aging; gene expression; healthy aging; human aging signatures; unhealthy aging.

Figure 1

Examples of age‐associated gene expression changes in GTEx subcutaneous fat. (a) Cartoon illustration of separating donors into the “tissue‐level healthy,” “common,” and “tissue‐level disease” cohorts. (b) Venn diagram shows the relationship among “common” aging genes, “healthy” aging genes, and DEGs calculated between the “tissue‐level healthy” cohort and the “tissue‐level disease” cohort in subcutaneous fat. (c, d, e) Scatter plots show three representative age‐associated gene expression patterns in three gene sets (from top to bottom: CAGs, CSAGs, or HSAGs). The red line and dots denote the regression line and gene expression levels for the “disease” cohort, the blue color is for the “healthy” cohort. Violin plots show the gene expression differences between “healthy” (blue) and “disease” (red) individuals for the corresponding genes displayed in the scatter plots

Figure 2

The relationship between age‐associated versus. disease‐associated genes. (a) The percentage of aging genes (blue: “healthy” aging; red: “common” aging) overlapped with disease DEGs and the percentage of disease DEGs that were not associated with age (gray bars). Disease DEGs include four disease signatures from prior work (insulin‐resistance, obesity, coronary heart disease [CHD], and chronic obstructive pulmonary disease [COPD]) and three disease DEGs based on GTEx analysis (simply labeled as DEG). The tissues plotted were subcutaneous fat (SF), tibial artery (TA), and lung. (b) The number of “healthy” aging genes whose direction was consistent (blue) or inconsistent (red) with the direction of gene regulations in 6 disease DEGs as in a (no common genes were found between COPD and “healthy” aging genes in lung). (c) An example of “healthy” aging gene in subcutaneous fat with the same direction of gene expression change as disease DEG. (d) An example of “healthy” aging gene in tibial artery with an opposite direction of gene expression change from disease DEG. The red line and dots denote the regression line and samples for the “disease” cohort, the blue line and dots are for the “healthy” cohort. Violin plots show the gene expression differences between “healthy” (blue) and “disease” (red) individuals

Figure 3

Function enrichment of the CAGs, CSAGs, and HSAGs. (a) GO terms and KEGG pathways enriched in the CAGs. (b) GO terms and KEGG pathways enriched in the CSAGs. (c) GO terms and KEGG pathways enriched in the HSAGs. The red bars denote upregulated genes with age, the blue bars represent downregulated genes with age

Figure 4

Enrichment of disease genes in three aging gene sets. (a, b) The enrichment between up/downregulated aging genes and complex disease genes in three aging gene sets (from left to right: CAGs, CSAGs or HSAGs) corresponding to four tissues (subcutaneous fat (SF); aorta artery (AA), tibial artery (TA), and Lung). −log₁₀ transformed p‐values were displayed in a color‐scale with more solid colors corresponding to more significant p‐values