Mitochondrial-targeted catalase is good for the old mouse proteome, but not for the young: 'reverse' antagonistic pleiotropy?

Abstract

Reactive oxygen species (ROS) are highly reactive oxygen-containing molecules associated with aging and a broad spectrum of pathologies. We have previously shown that transgenic expression of the antioxidant enzyme catalase targeted to the mitochondria (mCAT) in mice reduces ROS, attenuates age-related disease, and increases lifespan. However, it has been increasingly recognized that ROS also has beneficial roles in signaling, hormesis, stress response, and immunity. We therefore hypothesized that mCAT might be beneficial only when ROS approaches pathological levels in older age and might not be advantageous at a younger age when basal ROS is low. We analyzed abundance and turnover of the global proteome in hearts and livers of young (4 month) and old (20 month) mCAT and wild-type (WT) mice. In old hearts and livers of WT mice, protein half-lives were reduced compared to young, while in mCAT mice the reverse was observed; the longest half-lives were seen in old mCAT mice and the shortest in young mCAT. Protein abundance of old mCAT hearts recapitulated a more youthful proteomic expression profile (P-value < 0.01). However, young mCAT mice partially phenocopied the older wild-type proteome (P-value < 0.01). Age strongly interacts with mCAT, consistent with antagonistic pleiotropy in the reverse of the typical direction. These findings underscore the contrasting roles of ROS in young vs. old mice and indicate the need for better understanding of the interaction between dose and age in assessing the efficacy of therapeutic interventions in aging, including mitochondrial antioxidants.

Keywords: aging; antagonistic pleiotropy; catalase; mitochondria; protein turnover; reactive oxygen species.

Figure 1

Distributions of the changes in proteome half‐life in (A) heart and (B) liver tissues by treatment—OWT, YmCAT, and OmCAT—each relative to young WT controls. Dotted lines mark the location of a ratio equal to one (YWT).

Figure 2

Hierarchical heatmap of all cardiac proteins significantly changed in half‐life during aging and correlation plots of cardiac proteins that significantly changed with age. (A) Heatmap depicting all proteins with significantly altered (P‐value < 0.05) in the statistical comparison of YWT and OWT proteome half‐lives. The four columns correspond to the treatment groups used in this study: YWT, OWT, YmCAT, and OmCAT. Colors in the heatmap represent the half‐life of a protein relative to its mean half‐life (in days) across all treatment groups (difference from row mean). Correlations were made between OWT half‐lives vs. each treatment group—(B) YWT, (C) OmCAT, and (D) YmCAT. (E) A correlation of WT aging (OWT/YWT) vs. mCAT aging (OmCAT/YmCAT). * P‐value < 0.05, **P‐value < 0.001 for Spearman's correlation of the individual pathways in the regressions.

Figure 3

Heatmap and correlation plots of hepatic protein half‐lives that significantly changed with age. (A) Heatmap depicting all significantly changed (P‐value < 0.05) protein HLs during aging (OWT vs. YWT). The heatmap colors represent the half‐life of a protein relative to its mean half‐life (in days) across all treatment groups (difference from row mean). (B–D) Correlations of OWT half‐lives vs. (B) YWT, (C) OmCAT, and (D) YmCAT half‐lives. (E) A correlation of WT aging (OWT/YWT) vs. mCAT aging (OmCAT/YmCAT). ǂP‐value < 0.10, * P‐value < 0.05, **P‐value < 0.001 for Spearman's correlation of the individual pathways in the regressions.

Figure 4

Heatmap of all significant cardiac protein abundance changes with age and correlations of WT aging to mCAT aging. (A) The heatmap shows the direction and magnitude of abundance changes between the groups noted in the column labels. (B–D) Correlation plots of changes in HL of proteins in the top ten IPA pathways, (B) The effect of mCAT in young mice (YmCAT vs. YWT) positively correlates with WT aging (OWT vs. YWT, P‐value < 0.05). (C) The effect of mCAT in old mice (OmCAT vs. OWT) had little to no correlation with WT aging. (D) The effect of aging in mCAT (OmCAT vs. YmCAT) vs. WT aging (OWT vs. YWT) shows a significant negative correlation, with a correlation coefficient of −0.5. * P‐value < 0.05, **P‐value < 0.001 for Spearman's correlation of the individual pathway between the x‐ and y‐axis groups.

Figure 5

Reverse antagonistic pleiotropy in the proteome as a result of mCAT. We condensed proteomic changes into simple line plots depicting the trajectory of WT aging (black line) and mCAT aging (colored line). (See the Materials and methods for more detail.) (A) The heart proteome demonstrates reverse antagonistic pleiotropy: The YmCAT proteome more closely resembles old mice, while the OmCAT mouse is shifted toward a younger proteome compared to their OWT counterparts. (B) This effect was not seen in liver tissue, which (C) did not express mCAT protein, as measured by Western blotting and confirmed by qPCR (Fig S5). (D) Illustration depicting a model in which mCAT may be good in old mice, but not in young mice.

Figure 6

mCAT and aging alter markers of several protein turnover mechanisms. Western blotting was performed on various markers of turnover pathways and compared to the median global protein turnover rates determined by mass spectrometry in (A) heart and (B) liver. (C) Protein carbonylation increased with age and was attenuated in old mCAT in heart. (D) Liver tissue similarly increased in carbonylation with age, and this was attenuated in old mCAT mice. (E) Chaperone‐mediated autophagy, as measured by lamp2a, was increased in both young and old mCAT hearts. (F) Lamp2a was elevated in YmCAT as well as in OWT mouse liver. (G) Levels of polyubiquitinated proteins significantly increased with age in the heart, and this effect was attenuated in old mCAT. (H) Levels of polyubiquitinated proteins in the liver were increased by both mCAT and aging. * P‐value < 0.05.