Genome-wide transcript and protein analysis highlights the role of protein homeostasis in the aging mouse heart

Abstract

Investigation of the molecular mechanisms of aging in the human heart is challenging due to confounding factors, such as diet and medications, as well as limited access to tissues from healthy aging individuals. The laboratory mouse provides an ideal model to study aging in healthy individuals in a controlled environment. However, previous mouse studies have examined only a narrow range of the genetic variation that shapes individual differences during aging. Here, we analyze transcriptome and proteome data from 185 genetically diverse male and female mice at ages 6, 12 and 18 months to characterize molecular changes that occur in the aging heart. Transcripts and proteins reveal activation of pathways related to exocytosis and cellular transport with age, while processes involved in protein folding decrease with age. Additional changes are apparent only in the protein data including reduced fatty acid oxidation and increased autophagy. For proteins that form complexes, we see a decline in correlation between their component subunits with age, suggesting age-related loss of stoichiometry. The most affected complexes are themselves involved in protein homeostasis, which potentially contributes to a cycle of progressive breakdown in protein quality control with age. Our findings highlight the important role of post-transcriptional regulation in aging. In addition, we identify genetic loci that modulate age-related changes in protein homeostasis, suggesting that genetic variation can alter the molecular aging process.

Global and tissue-specific aging effects on murine proteomes

“…Despite heart tissue having far fewer significant sex effects than kidney or liver, consistent with our previous work 7 , they are still more consistent between proteins and transcripts than age effects.…”

“…The co-regulation of the overall proteosome complex, encompassing the 20S catalytic core, its 19S regulator, and the 11S regulator, is multifaceted. We have previously shown that abundance of individual components and sub-complexes of the proteasome are influenced by genetic variation 14 and age 7 in genetically diverse mouse populations. In this B6 sample population where genetic variation has been fixed (excluding spontaneous mutations specific to individuals), the sub-complex structure of the proteasome, most notably the 20S catalytic core and 19S regulator, is reflected in the pairwise correlations between complex members.…”

“…Protein abundance can vary between the sexes [13][14][15] and with age 6,7,16 . For each tissue, we characterized age and sex effects (Table S3) and declared differences to be significant based on false discovery rate (FDR) < 0.1 (Figures 2 and S2, respectively; Methods).…”

QTLViewer: An interactive webtool for genetic analysis in the Collaborative Cross and Diversity Outbred mouse populations

“…All QTLViewer projects contain at least one dataset such as a table of physiological traits, gene expression, or protein abundance data. Using the JAX Center for Aging Research DO data from the heart as an example (https://churchilllab.jax.org/qtlviewer/JAC/DOHeart) (Gerdes Gyuricza et al 2022) we find transcript and protein data available. Users can switch datasets by clicking the selection box arrow next to the text “Current Data Set” (Figure 3A).…”

“…The Collaborative Cross (CC) and Diversity Outbred (DO) mouse populations are related MPPs that have been used to genetically dissect complex traits, including obesity (Svenson et al 2012), bone strength (Al-Barghouthi et al 2021), short-term memory (Hsiao et al 2020), and arsenic response (French et al 2015). As recombinant populations, the CC and DO are well-suited for mapping quantitative trait loci (QTL), which can be powerful in studies of -omic traits, including gene expression, protein abundance, and chromatin accessibility (Abu Toamih Atamni et al 2018; Aylor et al 2011; Chick et al 2016; Gerdes Gyuricza et al 2022; Keele et al 2020; Keele et al 2021; Keller et al 2018; Takemon et al 2021).…”